feat: deduplicate code review findings across LLM passes

Group findings by file, line proximity, and normalized title keywords, keeping the highest-severity finding from each group and merging CWE info. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
fix: update rustls-webpki to 0.103.10 (RUSTSEC-2026-0049)
2026-03-25 21:55:52 +01:00 · 2026-03-25 17:08:18 +01:00 · 2026-03-25 16:48:08 +01:00 · 2026-03-25 16:32:34 +01:00 · 2026-03-20 22:51:39 +01:00 · 2026-03-18 23:41:03 +01:00
218 changed files with 2667 additions and 29544 deletions
@@ -1,23 +0,0 @@
-[advisories]
-ignore = [
-    # hickory-proto 0.25.x pulled in transitively via mongodb → hickory-resolver.
-    # MongoDB 3.x has not yet released with hickory-resolver 0.26.x, so we cannot
-    # upgrade past this without a mongodb release. Both are DNS-layer DoS vectors
-    # requiring a MITM/controlled DNS server against MongoDB's hostname resolution —
-    # not a realistic attack surface here. Revisit when mongodb bumps hickory.
-    "RUSTSEC-2026-0118", # NSEC3 loop, no fix available upstream
-    "RUSTSEC-2026-0119", # O(n²) name compression, fixed in hickory-proto >=0.26.1
-
-    # rmcp 0.16.0 — DNS rebinding in Streamable HTTP server transport (missing
-    # Host header validation). Patched in rmcp >= 1.4.0, which is a major API
-    # version jump from our pin; rmcp shipped 0.x → 1.x → 2.x in three months
-    # and the migration touches every tool handler + the auth middleware we
-    # just landed in #92. Threat model in our deployment: the MCP server is
-    # exposed at a public hostname (comp-mcp-dev.meghsakha.com) behind orca's
-    # TLS-terminating ingress with per-tenant bearer auth — the attack model
-    # (browser DNS-rebinding into localhost MCP server) doesn't directly apply.
-    # Defense-in-depth Host-header check is still a worthwhile follow-up.
-    # FOLLOW-UP: bump rmcp to 2.x in a dedicated PR (M7.3 follow-up, sized
-    # multi-hour due to API surface change).
-    "RUSTSEC-2026-0189",
-]
@@ -34,24 +34,6 @@ SCAN_SCHEDULE=0 0 */6 * * *
 CVE_MONITOR_SCHEDULE=0 0 0 * * *
 GIT_CLONE_BASE_PATH=/tmp/compliance-scanner/repos

-# Dynamic PLC testing — ephemeral soft-PLC provisioning (#183). Off unless
-# enabled; requires the agent container to have Docker access (socket mount).
-# When on, a PLC/SPS target with control logic but no reachable device gets its
-# logic instantiated on a throwaway OpenPLC, probed, then torn down.
-PLC_RUNTIME_ENABLED=0
-PLC_RUNTIME_IMAGE=registry.meghsakha.com/openplc:latest
-PLC_RUNTIME_NETWORK=certifai
-PLC_RUNTIME_MEMORY=512m
-PLC_RUNTIME_CPUS=0.5
-PLC_RUNTIME_MAX_LIFETIME_SECS=180
-PLC_RUNTIME_OPENPLC_USER=openplc
-PLC_RUNTIME_OPENPLC_PASSWORD=openplc
-
-# Werkbank runner API (/api/v1/werkbank/jobs/*, /api/v1/werkbank/artifacts/*).
-# When set, mounts the runner-facing queue + artifact endpoints behind this
-# bearer token; runners present the same token. Unset = endpoints not mounted.
-WERKBANK_RUNNER_TOKEN=
-
 # Dashboard
 DASHBOARD_PORT=8080
 AGENT_API_URL=http://localhost:3001
@@ -7,42 +7,17 @@ on:
  pull_request:

 env:
-  # registry + cosign creds via env, NOT inline ${{ }}: the Harbor robot
-  # username contains '$', which sh expands when interpolated into the
-  # script (robot$ci-push -> robot-push) => docker login unauthorized.
-  REGISTRY_USERNAME: ${{ secrets.REGISTRY_USERNAME }}
-  REGISTRY_PASSWORD: ${{ secrets.REGISTRY_PASSWORD }}
-  COSIGN_KEY: ${{ secrets.COSIGN_KEY }}
-  COSIGN_PASSWORD: ${{ secrets.COSIGN_PASSWORD }}
  CARGO_TERM_COLOR: always
  RUSTFLAGS: "-D warnings"
-  # Compile cache: sccache -> Hetzner S3 (breakpilot-sccache), runner-independent
-  # and persistent across CI runs (own key prefix). Reuses the shared cluster S3
-  # creds (same bucket as werkpilot). Requires repo secrets HETZNER_S3_ACCESS_KEY
-  # and HETZNER_S3_SECRET_KEY.
+  # sccache caches compilation artifacts within a job so that compiling
+  # both --features server and --features web shares common crate work.
  RUSTC_WRAPPER: /usr/local/bin/sccache
-  SCCACHE_BUCKET: breakpilot-sccache
-  SCCACHE_ENDPOINT: https://nbg1.your-objectstorage.com
-  SCCACHE_REGION: auto
-  SCCACHE_S3_USE_SSL: "true"
-  SCCACHE_S3_KEY_PREFIX: compliance-scanner
-  AWS_ACCESS_KEY_ID: ${{ secrets.HETZNER_S3_ACCESS_KEY }}
-  AWS_SECRET_ACCESS_KEY: ${{ secrets.HETZNER_S3_SECRET_KEY }}
-  # compliance-agent depends on tramiton-core via git; use the system git so the
-  # credential rewrite below (see "Configure git auth ...") is honored on fetch.
-  CARGO_NET_GIT_FETCH_WITH_CLI: "true"
-  # Throttle cargo so a ~670-crate concurrent download burst doesn't 429 the
-  # Kellnr mirror: fewer concurrent connections (HTTP/1.1) + more retries.
-  CARGO_NET_RETRY: "10"
-  CARGO_HTTP_MULTIPLEXING: "false"
+  SCCACHE_DIR: /tmp/sccache

-# Cancel superseded PR runs, but NEVER cancel main-branch runs — those build and
-# deploy per-service images, and cancelling one merge's deploy when the next
-# merge lands leaves a service un-deployed (as happened between two back-to-back
-# merges). So cancel-in-progress only for pull_request events.
+# Cancel in-progress runs for the same branch/PR
 concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
-  cancel-in-progress: ${{ github.event_name == 'pull_request' }}
+  cancel-in-progress: true

 jobs:
  # ---------------------------------------------------------------------------
@@ -61,44 +36,16 @@ jobs:
          git remote add origin "${GITHUB_SERVER_URL}/${GITHUB_REPOSITORY}.git"
          git fetch --depth=1 origin "${GITHUB_SHA}"
          git checkout FETCH_HEAD
-      # Resolve crates.io deps through the self-hosted Kellnr mirror (cached,
-      # crates.io-independent). Git deps (tramiton-core) are unaffected — source
-      # replacement only applies to crates.io-sourced crates.
-      - name: Use Kellnr crates.io mirror
-        run: |
-          : "${CARGO_HOME:=/usr/local/cargo}"
-          mkdir -p "$CARGO_HOME"
-          {
-            echo '[source.crates-io]'
-            echo 'replace-with = "kellnr"'
-            echo '[registries.kellnr]'
-            echo 'index = "sparse+https://crates.breakpilot.com/api/v1/cratesio/"'
-          } >> "$CARGO_HOME/config.toml"
-        env:
-          RUSTC_WRAPPER: ""
      - name: Install tools
        run: |
          rustup component add rustfmt clippy
-          curl -fsSL https://github.com/mozilla/sccache/releases/download/v0.10.0/sccache-v0.10.0-x86_64-unknown-linux-musl.tar.gz \
-            | tar xz --strip-components=1 -C /usr/local/bin/ sccache-v0.10.0-x86_64-unknown-linux-musl/sccache
+          curl -fsSL https://github.com/mozilla/sccache/releases/download/v0.9.1/sccache-v0.9.1-x86_64-unknown-linux-musl.tar.gz \
+            | tar xz --strip-components=1 -C /usr/local/bin/ sccache-v0.9.1-x86_64-unknown-linux-musl/sccache
          chmod +x /usr/local/bin/sccache
          cargo install cargo-audit --locked
        env:
          RUSTC_WRAPPER: ""

-      # compliance-agent has a git dependency on tramiton-core (a private repo on
-      # this Gitea instance). Rewrite its SSH URL to HTTPS + a PAT so the runner
-      # can fetch it. Requires the repo secret TRAMITON_FETCH_TOKEN (a Gitea PAT
-      # with read:repository, owned by a user with access to sharang/tramiton).
-      # (Honored on fetch because CARGO_NET_GIT_FETCH_WITH_CLI=true uses system git.)
-      - name: Configure git auth for private tramiton dependency
-        run: |
-          git config --global \
-            url."https://sharang:${{ secrets.TRAMITON_FETCH_TOKEN }}@git.breakpilot.com/".insteadOf \
-            "ssh://git@git.breakpilot.com:22222/"
-        env:
-          RUSTC_WRAPPER: ""
-
      # Format (no compilation needed)
      - name: Format
        run: cargo fmt --all --check
@@ -114,10 +61,6 @@ jobs:
        run: cargo clippy -p compliance-dashboard --features web --no-default-features -- -D warnings
      - name: Clippy (mcp)
        run: cargo clippy -p compliance-mcp -- -D warnings
-      - name: Clippy (werkbank-exec)
-        run: cargo clippy -p werkbank-exec -- -D warnings
-      - name: Clippy (control-map)
-        run: cargo clippy -p control-map -- -D warnings

      # Security audit
      - name: Security Audit
@@ -126,8 +69,8 @@ jobs:
          RUSTC_WRAPPER: ""

      # Tests (reuses compilation artifacts from clippy)
-      - name: Tests (core + agent + werkbank-exec + control-map)
-        run: cargo test -p compliance-core -p compliance-agent -p werkbank-exec -p control-map --lib
+      - name: Tests (core + agent)
+        run: cargo test -p compliance-core -p compliance-agent
      - name: Tests (dashboard server)
        run: cargo test -p compliance-dashboard --features server --no-default-features
      - name: Tests (dashboard web)
@@ -202,28 +145,13 @@ jobs:
    needs: [detect-changes]
    if: needs.detect-changes.outputs.agent == 'true'
    container:
-      image: docker:27-cli
+      image: alpine:latest
    steps:
-      - name: Build, push and trigger orca redeploy
-        env:
-          # PAT for fetching the private tramiton-core git dependency during the
-          # image build (injected as a BuildKit secret, never baked into a layer).
-          TRAMITON_FETCH_TOKEN: ${{ secrets.TRAMITON_FETCH_TOKEN }}
+      - name: Trigger Coolify deploy
        run: |
-          apk add --no-cache git curl openssl
-          git init && git remote add origin "${GITHUB_SERVER_URL}/${GITHUB_REPOSITORY}.git"
-          git fetch --depth=1 origin "${GITHUB_SHA}" && git checkout FETCH_HEAD
-          IMAGE=repo.breakpilot.com/certifai/compliance-agent
-          echo "$REGISTRY_PASSWORD" | docker login repo.breakpilot.com -u "$REGISTRY_USERNAME" --password-stdin
-          DOCKER_BUILDKIT=1 docker build --secret id=tramiton_token,env=TRAMITON_FETCH_TOKEN \
-            -f Dockerfile.agent -t "$IMAGE:latest" -t "$IMAGE:${GITHUB_SHA}" .
-          docker push "$IMAGE:latest" && docker push "$IMAGE:${GITHUB_SHA}"
-          { command -v cosign >/dev/null 2>&1 || curl -sSfLo /usr/local/bin/cosign https://github.com/sigstore/cosign/releases/download/v2.4.3/cosign-linux-amd64 || wget -qO /usr/local/bin/cosign https://github.com/sigstore/cosign/releases/download/v2.4.3/cosign-linux-amd64; } || echo "::warning::cosign fetch failed"
-          chmod +x /usr/local/bin/cosign 2>/dev/null || true
-          cosign sign --yes --key env://COSIGN_KEY "$IMAGE:latest" || echo "::warning::cosign failed"
-          PAYLOAD=$(printf '{"ref":"refs/heads/main","repository":{"full_name":"sharang/compliance-scanner-agent"},"head_commit":{"id":"%s","message":"deploy agent"}}' "${GITHUB_SHA}")
-          SIG=$(printf '%s' "$PAYLOAD" | openssl dgst -sha256 -hmac "${{ secrets.ORCA_WEBHOOK_SECRET }}" | awk '{print $2}')
-          RESP=$(curl -fsS -w "\nHTTP %{http_code}" -X POST "http://46.225.100.82:6880/api/v1/webhooks/github" -H "Content-Type: application/json" -H "X-Hub-Signature-256: sha256=$SIG" -d "$PAYLOAD"); echo "$RESP"
+          apk add --no-cache curl
+          curl -sf "${{ secrets.COOLIFY_WEBHOOK_AGENT }}" \
+            -H "Authorization: Bearer ${{ secrets.COOLIFY_TOKEN }}"

  deploy-dashboard:
    name: Deploy Dashboard
@@ -231,26 +159,13 @@ jobs:
    needs: [detect-changes]
    if: needs.detect-changes.outputs.dashboard == 'true'
    container:
-      image: docker:27-cli
+      image: alpine:latest
    steps:
-      - name: Build, push and trigger orca redeploy
-        env:
-          TRAMITON_FETCH_TOKEN: ${{ secrets.TRAMITON_FETCH_TOKEN }}
+      - name: Trigger Coolify deploy
        run: |
-          apk add --no-cache git curl openssl
-          git init && git remote add origin "${GITHUB_SERVER_URL}/${GITHUB_REPOSITORY}.git"
-          git fetch --depth=1 origin "${GITHUB_SHA}" && git checkout FETCH_HEAD
-          IMAGE=repo.breakpilot.com/certifai/compliance-dashboard
-          echo "$REGISTRY_PASSWORD" | docker login repo.breakpilot.com -u "$REGISTRY_USERNAME" --password-stdin
-          DOCKER_BUILDKIT=1 docker build --secret id=tramiton_token,env=TRAMITON_FETCH_TOKEN \
-            -f Dockerfile.dashboard -t "$IMAGE:latest" -t "$IMAGE:${GITHUB_SHA}" .
-          docker push "$IMAGE:latest" && docker push "$IMAGE:${GITHUB_SHA}"
-          { command -v cosign >/dev/null 2>&1 || curl -sSfLo /usr/local/bin/cosign https://github.com/sigstore/cosign/releases/download/v2.4.3/cosign-linux-amd64 || wget -qO /usr/local/bin/cosign https://github.com/sigstore/cosign/releases/download/v2.4.3/cosign-linux-amd64; } || echo "::warning::cosign fetch failed"
-          chmod +x /usr/local/bin/cosign 2>/dev/null || true
-          cosign sign --yes --key env://COSIGN_KEY "$IMAGE:latest" || echo "::warning::cosign failed"
-          PAYLOAD=$(printf '{"ref":"refs/heads/main","repository":{"full_name":"sharang/compliance-scanner-agent"},"head_commit":{"id":"%s","message":"deploy dashboard"}}' "${GITHUB_SHA}")
-          SIG=$(printf '%s' "$PAYLOAD" | openssl dgst -sha256 -hmac "${{ secrets.ORCA_WEBHOOK_SECRET }}" | awk '{print $2}')
-          RESP=$(curl -fsS -w "\nHTTP %{http_code}" -X POST "http://46.225.100.82:6880/api/v1/webhooks/github" -H "Content-Type: application/json" -H "X-Hub-Signature-256: sha256=$SIG" -d "$PAYLOAD"); echo "$RESP"
+          apk add --no-cache curl
+          curl -sf "${{ secrets.COOLIFY_WEBHOOK_DASHBOARD }}" \
+            -H "Authorization: Bearer ${{ secrets.COOLIFY_TOKEN }}"

  deploy-docs:
    name: Deploy Docs
@@ -258,23 +173,13 @@ jobs:
    needs: [detect-changes]
    if: needs.detect-changes.outputs.docs == 'true'
    container:
-      image: docker:27-cli
+      image: alpine:latest
    steps:
-      - name: Build, push and trigger orca redeploy
+      - name: Trigger Coolify deploy
        run: |
-          apk add --no-cache git curl openssl
-          git init && git remote add origin "${GITHUB_SERVER_URL}/${GITHUB_REPOSITORY}.git"
-          git fetch --depth=1 origin "${GITHUB_SHA}" && git checkout FETCH_HEAD
-          IMAGE=repo.breakpilot.com/certifai/compliance-docs
-          echo "$REGISTRY_PASSWORD" | docker login repo.breakpilot.com -u "$REGISTRY_USERNAME" --password-stdin
-          docker build -f Dockerfile.docs -t "$IMAGE:latest" -t "$IMAGE:${GITHUB_SHA}" .
-          docker push "$IMAGE:latest" && docker push "$IMAGE:${GITHUB_SHA}"
-          { command -v cosign >/dev/null 2>&1 || curl -sSfLo /usr/local/bin/cosign https://github.com/sigstore/cosign/releases/download/v2.4.3/cosign-linux-amd64 || wget -qO /usr/local/bin/cosign https://github.com/sigstore/cosign/releases/download/v2.4.3/cosign-linux-amd64; } || echo "::warning::cosign fetch failed"
-          chmod +x /usr/local/bin/cosign 2>/dev/null || true
-          cosign sign --yes --key env://COSIGN_KEY "$IMAGE:latest" || echo "::warning::cosign failed"
-          PAYLOAD=$(printf '{"ref":"refs/heads/main","repository":{"full_name":"sharang/compliance-scanner-agent"},"head_commit":{"id":"%s","message":"deploy docs"}}' "${GITHUB_SHA}")
-          SIG=$(printf '%s' "$PAYLOAD" | openssl dgst -sha256 -hmac "${{ secrets.ORCA_WEBHOOK_SECRET }}" | awk '{print $2}')
-          RESP=$(curl -fsS -w "\nHTTP %{http_code}" -X POST "http://46.225.100.82:6880/api/v1/webhooks/github" -H "Content-Type: application/json" -H "X-Hub-Signature-256: sha256=$SIG" -d "$PAYLOAD"); echo "$RESP"
+          apk add --no-cache curl
+          curl -sf "${{ secrets.COOLIFY_WEBHOOK_DOCS }}" \
+            -H "Authorization: Bearer ${{ secrets.COOLIFY_TOKEN }}"

  deploy-mcp:
    name: Deploy MCP
@@ -282,23 +187,10 @@ jobs:
    needs: [detect-changes]
    if: needs.detect-changes.outputs.mcp == 'true'
    container:
-      image: docker:27-cli
+      image: alpine:latest
    steps:
-      - name: Build, push and trigger orca redeploy
-        env:
-          TRAMITON_FETCH_TOKEN: ${{ secrets.TRAMITON_FETCH_TOKEN }}
+      - name: Trigger Coolify deploy
        run: |
-          apk add --no-cache git curl openssl
-          git init && git remote add origin "${GITHUB_SERVER_URL}/${GITHUB_REPOSITORY}.git"
-          git fetch --depth=1 origin "${GITHUB_SHA}" && git checkout FETCH_HEAD
-          IMAGE=repo.breakpilot.com/certifai/compliance-mcp
-          echo "$REGISTRY_PASSWORD" | docker login repo.breakpilot.com -u "$REGISTRY_USERNAME" --password-stdin
-          DOCKER_BUILDKIT=1 docker build --secret id=tramiton_token,env=TRAMITON_FETCH_TOKEN \
-            -f Dockerfile.mcp -t "$IMAGE:latest" -t "$IMAGE:${GITHUB_SHA}" .
-          docker push "$IMAGE:latest" && docker push "$IMAGE:${GITHUB_SHA}"
-          { command -v cosign >/dev/null 2>&1 || curl -sSfLo /usr/local/bin/cosign https://github.com/sigstore/cosign/releases/download/v2.4.3/cosign-linux-amd64 || wget -qO /usr/local/bin/cosign https://github.com/sigstore/cosign/releases/download/v2.4.3/cosign-linux-amd64; } || echo "::warning::cosign fetch failed"
-          chmod +x /usr/local/bin/cosign 2>/dev/null || true
-          cosign sign --yes --key env://COSIGN_KEY "$IMAGE:latest" || echo "::warning::cosign failed"
-          PAYLOAD=$(printf '{"ref":"refs/heads/main","repository":{"full_name":"sharang/compliance-scanner-agent"},"head_commit":{"id":"%s","message":"deploy mcp"}}' "${GITHUB_SHA}")
-          SIG=$(printf '%s' "$PAYLOAD" | openssl dgst -sha256 -hmac "${{ secrets.ORCA_WEBHOOK_SECRET }}" | awk '{print $2}')
-          RESP=$(curl -fsS -w "\nHTTP %{http_code}" -X POST "http://46.225.100.82:6880/api/v1/webhooks/github" -H "Content-Type: application/json" -H "X-Hub-Signature-256: sha256=$SIG" -d "$PAYLOAD"); echo "$RESP"
+          apk add --no-cache curl
+          curl -sf "${{ secrets.COOLIFY_WEBHOOK_MCP }}" \
+            -H "Authorization: Bearer ${{ secrets.COOLIFY_TOKEN }}"
@@ -1,52 +0,0 @@
-name: Nightly E2E Tests
-
-on:
-  schedule:
-    - cron: '0 3 * * *'  # 3 AM UTC daily
-  workflow_dispatch:       # Allow manual trigger
-
-env:
-  CARGO_TERM_COLOR: always
-  RUSTFLAGS: "-D warnings"
-  RUSTC_WRAPPER: /usr/local/bin/sccache
-  SCCACHE_DIR: /tmp/sccache
-  TEST_MONGODB_URI: "mongodb://root:example@mongo:27017/?authSource=admin"
-
-concurrency:
-  group: nightly-e2e
-  cancel-in-progress: true
-
-jobs:
-  e2e:
-    name: E2E Tests
-    runs-on: docker
-    container:
-      image: rust:1.94-bookworm
-    services:
-      mongo:
-        image: mongo:7
-        env:
-          MONGO_INITDB_ROOT_USERNAME: root
-          MONGO_INITDB_ROOT_PASSWORD: example
-    steps:
-      - name: Checkout
-        run: |
-          git init
-          git remote add origin "${GITHUB_SERVER_URL}/${GITHUB_REPOSITORY}.git"
-          git fetch --depth=1 origin "${GITHUB_SHA:-refs/heads/main}"
-          git checkout FETCH_HEAD
-
-      - name: Install sccache
-        run: |
-          curl -fsSL https://github.com/mozilla/sccache/releases/download/v0.9.1/sccache-v0.9.1-x86_64-unknown-linux-musl.tar.gz \
-            | tar xz --strip-components=1 -C /usr/local/bin/ sccache-v0.9.1-x86_64-unknown-linux-musl/sccache
-          chmod +x /usr/local/bin/sccache
-        env:
-          RUSTC_WRAPPER: ""
-
-      - name: Run E2E tests
-        run: cargo test -p compliance-agent --test e2e -- --test-threads=4
-
-      - name: Show sccache stats
-        run: sccache --show-stats
-        if: always()
@@ -666,7 +666,6 @@ dependencies = [
 "compliance-core",
 "compliance-dast",
 "compliance-graph",
- "control-map",
 "dashmap",
 "dotenvy",
 "futures-core",
@@ -677,10 +676,8 @@ dependencies = [
 "jsonwebtoken",
 "mongodb",
 "octocrab",
- "rand 0.9.2",
 "regex",
 "reqwest",
- "roxmltree",
 "secrecy",
 "serde",
 "serde_json",
@@ -690,17 +687,12 @@ dependencies = [
 "tokio-cron-scheduler",
 "tokio-stream",
 "tokio-tungstenite 0.26.2",
- "tower",
 "tower-http",
 "tracing",
 "tracing-subscriber",
- "tramiton-core",
- "tramiton-repro",
- "tramiton-sbom",
 "urlencoding",
 "uuid",
 "walkdir",
- "werkbank-exec",
 "zip",
 ]

@@ -708,24 +700,19 @@ dependencies = [
 name = "compliance-core"
 version = "0.1.0"
 dependencies = [
- "axum",
 "bson",
 "chrono",
 "hex",
- "jsonwebtoken",
 "mongodb",
 "opentelemetry",
 "opentelemetry-appender-tracing",
 "opentelemetry-otlp",
 "opentelemetry_sdk",
- "reqwest",
 "secrecy",
 "serde",
 "serde_json",
 "sha2",
 "thiserror 2.0.18",
- "tokio",
- "toml",
 "tracing",
 "tracing-opentelemetry",
 "tracing-subscriber",
@@ -826,15 +813,12 @@ dependencies = [
 "bson",
 "chrono",
 "compliance-core",
- "dashmap",
 "dotenvy",
- "hex",
 "mongodb",
 "rmcp",
 "schemars 1.2.1",
 "serde",
 "serde_json",
- "sha2",
 "thiserror 2.0.18",
 "tokio",
 "tower-http",
@@ -842,20 +826,6 @@ dependencies = [
 "tracing-subscriber",
 ]

-[[package]]
-name = "compliance-smoke"
-version = "0.1.0"
-dependencies = [
- "axum",
- "compliance-core",
- "reqwest",
- "serde",
- "serde_json",
- "tokio",
- "tracing",
- "tracing-subscriber",
-]
-
 [[package]]
 name = "console_error_panic_hook"
 version = "0.1.7"
@@ -969,15 +939,6 @@ dependencies = [
 "charset",
 ]

-[[package]]
-name = "control-map"
-version = "0.1.0"
-dependencies = [
- "serde",
- "serde_json",
- "thiserror 2.0.18",
-]
-
 [[package]]
 name = "convert_case"
 version = "0.8.0"
@@ -1134,9 +1095,9 @@ dependencies = [

 [[package]]
 name = "crossbeam-epoch"
-version = "0.9.20"
+version = "0.9.18"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2d6914041f254d6e9176c01941b21115dcfb7089e55135a35411081bd106ef3f"
+checksum = "5b82ac4a3c2ca9c3460964f020e1402edd5753411d7737aa39c3714ad1b5420e"
 dependencies = [
 "crossbeam-utils",
 ]
@@ -3563,9 +3524,9 @@ checksum = "224484c5d09285a7b8cb0a0c117e847ebd14cb6e4470ecf68cdb89c503b0edb9"

 [[package]]
 name = "mongodb"
-version = "3.6.0"
+version = "3.5.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1ef2c933617431ad0246fb5b43c425ebdae18c7f7259c87de0726d93b0e7e91b"
+checksum = "803dd859e8afa084c255a8effd8000ff86f7c8076a50cd6d8c99e8f3496f75c2"
 dependencies = [
 "base64",
 "bitflags",
@@ -3609,9 +3570,9 @@ dependencies = [

 [[package]]
 name = "mongodb-internal-macros"
-version = "3.6.0"
+version = "3.5.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9e5758dc828eb2d02ec30563cba365609d56ddd833190b192beaee2b475a7bb3"
+checksum = "a973ef3dd3dbc6f6e65bbdecfd9ec5e781b9e7493b0f369a7c62e35d8e5ae2c8"
 dependencies = [
 "macro_magic",
 "proc-macro2",
@@ -3782,15 +3743,6 @@ dependencies = [
 "syn",
 ]

-[[package]]
-name = "object"
-version = "0.36.7"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "62948e14d923ea95ea2c7c86c71013138b66525b86bdc08d2dcc262bdb497b87"
-dependencies = [
- "memchr",
-]
-
 [[package]]
 name = "octocrab"
 version = "0.44.1"
@@ -4222,7 +4174,7 @@ version = "3.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "219cb19e96be00ab2e37d6e299658a0cfa83e52429179969b0f0121b4ac46983"
 dependencies = [
- "toml_edit 0.23.10+spec-1.0.0",
+ "toml_edit",
 ]

 [[package]]
@@ -4307,9 +4259,9 @@ dependencies = [

 [[package]]
 name = "quinn-proto"
-version = "0.11.15"
+version = "0.11.14"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4fcb935c5bec503c2f0e306bdd3e58bb9029dcb14fa8d9ac76e3a5256ac0763e"
+checksum = "434b42fec591c96ef50e21e886936e66d3cc3f737104fdb9b737c40ffb94c098"
 dependencies = [
 "bytes",
 "getrandom 0.3.4",
@@ -4641,12 +4593,6 @@ dependencies = [
 "syn",
 ]

-[[package]]
-name = "roxmltree"
-version = "0.20.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6c20b6793b5c2fa6553b250154b78d6d0db37e72700ae35fad9387a46f487c97"
-
 [[package]]
 name = "rust-stemmers"
 version = "1.2.0"
@@ -4753,9 +4699,9 @@ dependencies = [

 [[package]]
 name = "rustls-webpki"
-version = "0.103.13"
+version = "0.103.10"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "61c429a8649f110dddef65e2a5ad240f747e85f7758a6bccc7e5777bd33f756e"
+checksum = "df33b2b81ac578cabaf06b89b0631153a3f416b0a886e8a7a1707fb51abbd1ef"
 dependencies = [
 "ring",
 "rustls-pki-types",
@@ -5027,15 +4973,6 @@ dependencies = [
 "syn",
 ]

-[[package]]
-name = "serde_spanned"
-version = "0.6.9"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bf41e0cfaf7226dca15e8197172c295a782857fcb97fad1808a166870dee75a3"
-dependencies = [
- "serde",
-]
-
 [[package]]
 name = "serde_urlencoded"
 version = "0.7.1"
@@ -5099,12 +5036,6 @@ dependencies = [
 "digest",
 ]

-[[package]]
-name = "sha1_smol"
-version = "1.0.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "bbfa15b3dddfee50a0fff136974b3e1bde555604ba463834a7eb7deb6417705d"
-
 [[package]]
 name = "sha2"
 version = "0.10.9"
@@ -5240,7 +5171,7 @@ version = "0.8.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "c1c97747dbf44bb1ca44a561ece23508e99cb592e862f22222dcf42f51d1e451"
 dependencies = [
- "heck 0.5.0",
+ "heck 0.4.1",
 "proc-macro2",
 "quote",
 "syn",
@@ -5856,27 +5787,6 @@ dependencies = [
 "tokio",
 ]

-[[package]]
-name = "toml"
-version = "0.8.23"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "dc1beb996b9d83529a9e75c17a1686767d148d70663143c7854d8b4a09ced362"
-dependencies = [
- "serde",
- "serde_spanned",
- "toml_datetime 0.6.11",
- "toml_edit 0.22.27",
-]
-
-[[package]]
-name = "toml_datetime"
-version = "0.6.11"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "22cddaf88f4fbc13c51aebbf5f8eceb5c7c5a9da2ac40a13519eb5b0a0e8f11c"
-dependencies = [
- "serde",
-]
-
 [[package]]
 name = "toml_datetime"
 version = "0.7.5+spec-1.1.0"
@@ -5886,20 +5796,6 @@ dependencies = [
 "serde_core",
 ]

-[[package]]
-name = "toml_edit"
-version = "0.22.27"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "41fe8c660ae4257887cf66394862d21dbca4a6ddd26f04a3560410406a2f819a"
-dependencies = [
- "indexmap 2.13.0",
- "serde",
- "serde_spanned",
- "toml_datetime 0.6.11",
- "toml_write",
- "winnow",
-]
-
 [[package]]
 name = "toml_edit"
 version = "0.23.10+spec-1.0.0"
@@ -5907,7 +5803,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "84c8b9f757e028cee9fa244aea147aab2a9ec09d5325a9b01e0a49730c2b5269"
 dependencies = [
 "indexmap 2.13.0",
- "toml_datetime 0.7.5+spec-1.1.0",
+ "toml_datetime",
 "toml_parser",
 "winnow",
 ]
@@ -5921,12 +5817,6 @@ dependencies = [
 "winnow",
 ]

-[[package]]
-name = "toml_write"
-version = "0.1.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5d99f8c9a7727884afe522e9bd5edbfc91a3312b36a77b5fb8926e4c31a41801"
-
 [[package]]
 name = "tonic"
 version = "0.12.3"
@@ -6173,46 +6063,6 @@ dependencies = [
 "wasm-bindgen",
 ]

-[[package]]
-name = "tramiton-core"
-version = "0.4.1"
-source = "git+ssh://git@gitea.meghsakha.com:22222/sharang/tramiton.git?tag=v0.4.1#ae4fc1376279f9edb9882605b20877335e7ba8ba"
-dependencies = [
- "serde",
- "tempfile",
- "thiserror 1.0.69",
- "toml",
- "walkdir",
-]
-
-[[package]]
-name = "tramiton-repro"
-version = "0.4.1"
-source = "git+ssh://git@gitea.meghsakha.com:22222/sharang/tramiton.git?tag=v0.4.1#ae4fc1376279f9edb9882605b20877335e7ba8ba"
-dependencies = [
- "serde",
- "serde_json",
- "sha2",
- "tempfile",
- "thiserror 1.0.69",
- "toml",
- "tramiton-core",
- "walkdir",
-]
-
-[[package]]
-name = "tramiton-sbom"
-version = "0.4.1"
-source = "git+ssh://git@gitea.meghsakha.com:22222/sharang/tramiton.git?tag=v0.4.1#ae4fc1376279f9edb9882605b20877335e7ba8ba"
-dependencies = [
- "object",
- "serde",
- "serde_json",
- "sha2",
- "tramiton-core",
- "tramiton-repro",
-]
-
 [[package]]
 name = "tree-sitter"
 version = "0.24.7"
@@ -6488,7 +6338,6 @@ dependencies = [
 "getrandom 0.4.1",
 "js-sys",
 "serde_core",
- "sha1_smol",
 "wasm-bindgen",
 ]

@@ -6732,25 +6581,6 @@ dependencies = [
 "rustls-pki-types",
 ]

-[[package]]
-name = "werkbank-exec"
-version = "0.1.0"
-dependencies = [
- "compliance-core",
- "compliance-dast",
- "futures-util",
- "hex",
- "regex",
- "reqwest",
- "secrecy",
- "sha2",
- "thiserror 2.0.18",
- "tokio",
- "tracing",
- "uuid",
- "walkdir",
-]
-
 [[package]]
 name = "which"
 version = "6.0.3"
@@ -6,9 +6,6 @@ members = [
    "compliance-graph",
    "compliance-dast",
    "compliance-mcp",
-    "compliance-smoke",
-    "werkbank-exec",
-    "control-map",
 ]
 resolver = "2"

@@ -18,7 +15,6 @@ expect_used = "deny"

 [workspace.dependencies]
 compliance-core = { path = "compliance-core", default-features = false }
-control-map = { path = "control-map" }
 serde = { version = "1", features = ["derive"] }
 serde_json = "1"
 tokio = { version = "1", features = ["full"] }
@@ -26,16 +22,14 @@ tracing = "0.1"
 tracing-subscriber = { version = "0.3", features = ["env-filter"] }
 chrono = { version = "0.4", features = ["serde"] }
 mongodb = { version = "3", features = ["rustls-tls", "compat-3-0-0"] }
-reqwest = { version = "0.12", features = ["json", "rustls-tls", "multipart", "cookies"], default-features = false }
+reqwest = { version = "0.12", features = ["json", "rustls-tls"], default-features = false }
 thiserror = "2"
 sha2 = "0.10"
 hex = "0.4"
-uuid = { version = "1", features = ["v4", "v5", "serde"] }
+uuid = { version = "1", features = ["v4", "serde"] }
 secrecy = { version = "0.10", features = ["serde"] }
 regex = "1"
 zip = { version = "2", features = ["aes-crypto", "deflate"] }
 dashmap = "6"
 tokio-stream = { version = "0.1", features = ["sync"] }
 aes-gcm = "0.10"
-rand = "0.9"
-base64 = "0.22"
@@ -2,22 +2,7 @@ FROM rust:1.94-bookworm AS builder

 WORKDIR /app
 COPY . .
-# compliance-agent depends on the private tramiton-core git repo. Authenticate
-# the fetch with a PAT passed as a BuildKit secret (never baked into a layer).
-# Build with: DOCKER_BUILDKIT=1 docker build --secret id=tramiton_token,env=TRAMITON_FETCH_TOKEN ...
-RUN --mount=type=secret,id=tramiton_token \
-    if [ -s /run/secrets/tramiton_token ]; then \
-      git config --global \
-        url."https://sharang:$(cat /run/secrets/tramiton_token)@git.breakpilot.com/".insteadOf \
-        "ssh://git@git.breakpilot.com:22222/"; \
-    fi && \
-    CARGO_NET_GIT_FETCH_WITH_CLI=true cargo build --release -p compliance-agent
-
-# A throwaway stage that packs a real nix store (store paths + the validity DB)
-# into a compressed bootstrap tarball. Only the tarball is copied into the final
-# image, so we don't carry a raw /nix copy layer.
-FROM nixos/nix:latest AS nixseed
-RUN tar -C / -czf /nix-bootstrap.tar.gz nix
+RUN cargo build --release -p compliance-agent

 FROM debian:bookworm-slim
 RUN apt-get update && apt-get install -y ca-certificates libssl3 git curl python3 python3-pip npm golang-go php-cli && rm -rf /var/lib/apt/lists/*
@@ -46,41 +31,11 @@ RUN pip3 install --break-system-packages semgrep
 # Install ruff for Python linting
 RUN pip3 install --break-system-packages ruff

-# Real nix for the tramiton reproducible-build firmware SBOM.
-#
-# nix-portable's proot fallback can't run here: user namespaces are blocked by
-# the container's default seccomp/apparmor profile, and orca exposes no way to
-# relax it. So ship a *real* nix and disable its build sandbox
-# (`sandbox = false`) — a plain gcc/make firmware build needs no user namespace,
-# so it runs fine under the locked-down profile with no proot involved.
-#
-# The store is shipped as a bootstrap tarball and seeded onto /nix at first
-# start (see docker/agent-entrypoint.sh), so a persistent /nix volume survives
-# redeploys. A missing/broken nix just falls back to the analysis-only SBOM.
-COPY --from=nixseed /nix-bootstrap.tar.gz /opt/nix-bootstrap.tar.gz
-ENV PATH="/nix/var/nix/profiles/default/bin:${PATH}"
-RUN mkdir -p /etc/nix && printf '%s\n' \
-      'experimental-features = nix-command flakes' \
-      'sandbox = false' \
-      'build-users-group =' \
-      'substituters = https://cache.nixos.org' \
-      'trusted-public-keys = cache.nixos.org-1:6NCHdD59X431o0gWypbMrAURkbJ16ZPMQFGspcDShjY=' \
-      > /etc/nix/nix.conf
-
 COPY --from=builder /app/target/release/compliance-agent /usr/local/bin/compliance-agent
-COPY docker/agent-entrypoint.sh /usr/local/bin/agent-entrypoint.sh
-RUN chmod +x /usr/local/bin/agent-entrypoint.sh
-
-# Copy documentation for the help chat assistant
-COPY --from=builder /app/README.md /app/README.md
-COPY --from=builder /app/docs /app/docs
-ENV HELP_DOCS_PATH=/app

 # Ensure SSH key directory exists
 RUN mkdir -p /data/compliance-scanner/ssh

 EXPOSE 3001 3002

-# Seeds /nix (fresh volume) from the bootstrap tarball, then runs the agent.
-ENTRYPOINT ["/usr/local/bin/agent-entrypoint.sh"]
-
+ENTRYPOINT ["compliance-agent"]
@@ -1,22 +1,13 @@
 FROM rust:1.94-bookworm AS builder

-RUN cargo install dioxus-cli --version 0.7.3 --locked
+RUN cargo install dioxus-cli --version 0.7.3

 ARG DOCS_URL=/docs

 WORKDIR /app
 COPY . .
 ENV DOCS_URL=${DOCS_URL}
-# compliance-agent (a workspace member) depends on the private tramiton-core git
-# repo, so the workspace resolve needs it even to build the dashboard.
-# Authenticate the fetch with a PAT passed as a BuildKit secret.
-RUN --mount=type=secret,id=tramiton_token \
-    if [ -s /run/secrets/tramiton_token ]; then \
-      git config --global \
-        url."https://sharang:$(cat /run/secrets/tramiton_token)@git.breakpilot.com/".insteadOf \
-        "ssh://git@git.breakpilot.com:22222/"; \
-    fi && \
-    CARGO_NET_GIT_FETCH_WITH_CLI=true dx build --release --package compliance-dashboard
+RUN dx build --release --package compliance-dashboard

 FROM debian:bookworm-slim
 RUN apt-get update && apt-get install -y ca-certificates libssl3 && rm -rf /var/lib/apt/lists/*
@@ -29,4 +20,3 @@ ENV IP=0.0.0.0
 EXPOSE 8080

 ENTRYPOINT ["./compliance-dashboard"]
-
@@ -12,4 +12,3 @@ RUN rm /etc/nginx/conf.d/default.conf
 COPY docs/nginx.conf /etc/nginx/conf.d/default.conf
 COPY --from=builder /app/.vitepress/dist /usr/share/nginx/html
 EXPOSE 80
-
@@ -2,16 +2,7 @@ FROM rust:1.94-bookworm AS builder

 WORKDIR /app
 COPY . .
-# compliance-agent (a workspace member) depends on the private tramiton-core git
-# repo, so the workspace resolve needs it even to build the mcp binary.
-# Authenticate the fetch with a PAT passed as a BuildKit secret.
-RUN --mount=type=secret,id=tramiton_token \
-    if [ -s /run/secrets/tramiton_token ]; then \
-      git config --global \
-        url."https://sharang:$(cat /run/secrets/tramiton_token)@git.breakpilot.com/".insteadOf \
-        "ssh://git@git.breakpilot.com:22222/"; \
-    fi && \
-    CARGO_NET_GIT_FETCH_WITH_CLI=true cargo build --release -p compliance-mcp
+RUN cargo build --release -p compliance-mcp

 FROM debian:bookworm-slim
 RUN apt-get update && apt-get install -y ca-certificates libssl3 && rm -rf /var/lib/apt/lists/*
@@ -23,4 +14,3 @@ EXPOSE 8090
 ENV MCP_PORT=8090

 ENTRYPOINT ["compliance-mcp"]
-
@@ -28,9 +28,9 @@

 ## About

-Compliance Scanner is an autonomous agent that continuously monitors git repositories for security vulnerabilities, GDPR/OAuth compliance patterns, and dependency risks. It creates issues in external trackers (GitHub/GitLab/Jira/Gitea) with evidence and remediation suggestions, reviews pull requests with multi-pass LLM analysis, runs autonomous penetration tests, and exposes a Dioxus-based dashboard for visualization.
+Compliance Scanner is an autonomous agent that continuously monitors git repositories for security vulnerabilities, GDPR/OAuth compliance patterns, and dependency risks. It creates issues in external trackers (GitHub/GitLab/Jira) with evidence and remediation suggestions, reviews pull requests, and exposes a Dioxus-based dashboard for visualization.

-> **How it works:** The agent runs as a lazy daemon -- it only scans when new commits are detected, triggered by cron schedules or webhooks. LLM-powered triage filters out false positives and generates actionable remediation with multi-language awareness.
+> **How it works:** The agent runs as a lazy daemon -- it only scans when new commits are detected, triggered by cron schedules or webhooks. LLM-powered triage filters out false positives and generates actionable remediation.

 ## Features

@@ -41,38 +41,31 @@ Compliance Scanner is an autonomous agent that continuously monitors git reposit
 | **CVE Monitoring** | OSV.dev batch queries, NVD CVSS enrichment, SearXNG context |
 | **GDPR Patterns** | Detect PII logging, missing consent, hardcoded retention, missing deletion |
 | **OAuth Patterns** | Detect implicit grant, missing PKCE, token in localStorage, token in URLs |
-| **LLM Triage** | Multi-language-aware confidence scoring (Rust, Python, Go, Java, Ruby, PHP, C++) |
-| **Issue Creation** | Auto-create issues in GitHub, GitLab, Jira, or Gitea with dedup via fingerprints |
-| **PR Reviews** | Multi-pass security review (logic, security, convention, complexity) with dedup |
-| **DAST Scanning** | Black-box security testing with endpoint discovery and parameter fuzzing |
-| **AI Pentesting** | Autonomous LLM-orchestrated penetration testing with encrypted reports |
-| **Code Graph** | Interactive code knowledge graph with impact analysis |
-| **AI Chat (RAG)** | Natural language Q&A grounded in repository source code |
-| **Help Assistant** | Documentation-grounded help chat accessible from every dashboard page |
-| **MCP Server** | Expose live security data to Claude, Cursor, and other AI tools |
-| **Dashboard** | Fullstack Dioxus UI with findings, SBOM, issues, DAST, pentest, and graph |
-| **Webhooks** | GitHub, GitLab, and Gitea webhook receivers for push/PR events |
-| **Finding Dedup** | SHA-256 fingerprint dedup for SAST, CWE-based dedup for DAST findings |
+| **LLM Triage** | Confidence scoring via LiteLLM to filter false positives |
+| **Issue Creation** | Auto-create issues in GitHub, GitLab, or Jira with code evidence |
+| **PR Reviews** | Post security review comments on pull requests |
+| **Dashboard** | Fullstack Dioxus UI with findings, SBOM, issues, and statistics |
+| **Webhooks** | GitHub (HMAC-SHA256) and GitLab webhook receivers for push/PR events |

 ## Architecture

 ```
-┌──────────────────────────────────────────────────────────────────────────┐
-│                          Cargo Workspace                                 │
-├──────────────┬──────────────────┬──────────────┬──────────┬─────────────┤
-│ compliance-  │ compliance-      │ compliance-  │ complian-│ compliance- │
-│ core (lib)   │ agent (bin)      │ dashboard    │ ce-graph │ mcp (bin)   │
-│              │                  │ (bin)        │ (lib)    │             │
-│ Models       │ Scan Pipeline    │ Dioxus 0.7   │ Tree-    │ MCP Server  │
-│ Traits       │ LLM Client      │ Fullstack UI │ sitter   │ Live data   │
-│ Config       │ Issue Trackers   │ Help Chat    │ Graph    │ for AI      │
-│ Errors       │ Pentest Engine   │ Server Fns   │ Embedds  │ tools       │
-│              │ DAST Tools       │              │ RAG      │             │
-│              │ REST API         │              │          │             │
-│              │ Webhooks         │              │          │             │
-└──────────────┴──────────────────┴──────────────┴──────────┴─────────────┘
-                                 │
-                            MongoDB (shared)
+┌─────────────────────────────────────────────────────────────┐
+│                    Cargo Workspace                           │
+├──────────────┬──────────────────┬───────────────────────────┤
+│ compliance-  │ compliance-      │ compliance-               │
+│ core         │ agent            │ dashboard                 │
+│ (lib)        │ (bin)            │ (bin, Dioxus 0.7.3)       │
+│              │                  │                           │
+│ Models       │ Scan Pipeline    │ Fullstack Web UI          │
+│ Traits       │ LLM Client      │ Server Functions           │
+│ Config       │ Issue Trackers   │ Charts + Tables           │
+│ Errors       │ Scheduler        │ Settings Page             │
+│              │ REST API         │                           │
+│              │ Webhooks         │                           │
+└──────────────┴──────────────────┴───────────────────────────┘
+                        │
+                   MongoDB (shared)
 ```

 ## Scan Pipeline (7 Stages)
@@ -91,16 +84,11 @@ Compliance Scanner is an autonomous agent that continuously monitors git reposit
 |-------|-----------|
 | Shared Library | `compliance-core` -- models, traits, config |
 | Agent | Axum REST API, git2, tokio-cron-scheduler, Semgrep, Syft |
-| Dashboard | Dioxus 0.7.3 fullstack, Tailwind CSS 4 |
-| Code Graph | `compliance-graph` -- tree-sitter parsing, embeddings, RAG |
-| MCP Server | `compliance-mcp` -- Model Context Protocol for AI tools |
-| DAST | `compliance-dast` -- dynamic application security testing |
+| Dashboard | Dioxus 0.7.3 fullstack, Tailwind CSS |
 | Database | MongoDB with typed collections |
-| LLM | LiteLLM (OpenAI-compatible API for chat, triage, embeddings) |
-| Issue Trackers | GitHub (octocrab), GitLab (REST v4), Jira (REST v3), Gitea |
+| LLM | LiteLLM (OpenAI-compatible API) |
+| Issue Trackers | GitHub (octocrab), GitLab (REST v4), Jira (REST v3) |
 | CVE Sources | OSV.dev, NVD, SearXNG |
-| Auth | Keycloak (OAuth2/PKCE, SSO) |
-| Browser Automation | Chromium (headless, for pentesting and PDF generation) |

 ## Getting Started

@@ -163,35 +151,20 @@ The agent exposes a REST API on port 3001:
 | `GET` | `/api/v1/sbom` | List dependencies |
 | `GET` | `/api/v1/issues` | List cross-tracker issues |
 | `GET` | `/api/v1/scan-runs` | Scan execution history |
-| `GET` | `/api/v1/graph/:repo_id` | Code knowledge graph |
-| `POST` | `/api/v1/graph/:repo_id/build` | Trigger graph build |
-| `GET` | `/api/v1/dast/targets` | List DAST targets |
-| `POST` | `/api/v1/dast/targets` | Add DAST target |
-| `GET` | `/api/v1/dast/findings` | List DAST findings |
-| `POST` | `/api/v1/chat/:repo_id` | RAG-powered code chat |
-| `POST` | `/api/v1/help/chat` | Documentation-grounded help chat |
-| `POST` | `/api/v1/pentest/sessions` | Create pentest session |
-| `POST` | `/api/v1/pentest/sessions/:id/export` | Export encrypted pentest report |
 | `POST` | `/webhook/github` | GitHub webhook (HMAC-SHA256) |
 | `POST` | `/webhook/gitlab` | GitLab webhook (token verify) |
-| `POST` | `/webhook/gitea` | Gitea webhook |

 ## Dashboard Pages

 | Page | Description |
 |------|-------------|
-| **Overview** | Stat cards, severity distribution, AI chat cards, MCP status |
-| **Repositories** | Add/manage tracked repos, trigger scans, webhook config |
-| **Findings** | Filterable table by severity, type, status, scanner |
+| **Overview** | Stat cards, severity distribution chart |
+| **Repositories** | Add/manage tracked repos, trigger scans |
+| **Findings** | Filterable table by severity, type, status |
 | **Finding Detail** | Code evidence, remediation, suggested fix, linked issue |
-| **SBOM** | Dependency inventory with vulnerability badges, license summary |
-| **Issues** | Cross-tracker view (GitHub + GitLab + Jira + Gitea) |
-| **Code Graph** | Interactive architecture visualization, impact analysis |
-| **AI Chat** | RAG-powered Q&A about repository code |
-| **DAST** | Dynamic scanning targets, findings, and scan history |
-| **Pentest** | AI-driven pentest sessions, attack chain visualization |
-| **MCP Servers** | Model Context Protocol server management |
-| **Help Chat** | Floating assistant (available on every page) for product Q&A |
+| **SBOM** | Dependency inventory with vulnerability badges |
+| **Issues** | Cross-tracker view (GitHub + GitLab + Jira) |
+| **Settings** | Configure LiteLLM, tracker tokens, SearXNG URL |

 ## Project Structure

@@ -200,24 +173,19 @@ compliance-scanner/
 ├── compliance-core/        Shared library (models, traits, config, errors)
 ├── compliance-agent/       Agent daemon (pipeline, LLM, trackers, API, webhooks)
 │   └── src/
-│       ├── pipeline/       7-stage scan pipeline, dedup, PR reviews, code review
-│       ├── llm/            LiteLLM client, triage, descriptions, fixes, review prompts
-│       ├── trackers/       GitHub, GitLab, Jira, Gitea integrations
-│       ├── pentest/        AI-driven pentest orchestrator, tools, reports
-│       ├── rag/            RAG pipeline, chunking, embedding
-│       ├── api/            REST API (Axum), help chat
-│       └── webhooks/       GitHub, GitLab, Gitea webhook receivers
+│       ├── pipeline/       7-stage scan pipeline
+│       ├── llm/            LiteLLM client, triage, descriptions, fixes, PR review
+│       ├── trackers/       GitHub, GitLab, Jira integrations
+│       ├── api/            REST API (Axum)
+│       └── webhooks/       GitHub + GitLab webhook receivers
 ├── compliance-dashboard/   Dioxus fullstack dashboard
 │   └── src/
-│       ├── components/     Reusable UI (sidebar, help chat, attack chain, etc.)
-│       ├── infrastructure/ Server functions, DB, config, auth
-│       └── pages/          Full page views (overview, DAST, pentest, graph, etc.)
-├── compliance-graph/       Code knowledge graph (tree-sitter, embeddings, RAG)
-├── compliance-dast/        Dynamic application security testing
-├── compliance-mcp/         Model Context Protocol server
-├── docs/                   VitePress documentation site
+│       ├── components/     Reusable UI components
+│       ├── infrastructure/ Server functions, DB, config
+│       └── pages/          Full page views
 ├── assets/                 Static assets (CSS, icons)
-└── styles/                 Tailwind input stylesheet
+├── styles/                 Tailwind input stylesheet
+└── bin/                    Dashboard binary entrypoint
 ```

 ## External Services
@@ -225,12 +193,10 @@ compliance-scanner/
 | Service | Purpose | Default URL |
 |---------|---------|-------------|
 | MongoDB | Persistence | `mongodb://localhost:27017` |
-| LiteLLM | LLM proxy (chat, triage, embeddings) | `http://localhost:4000` |
+| LiteLLM | LLM proxy for triage and generation | `http://localhost:4000` |
 | SearXNG | CVE context search | `http://localhost:8888` |
-| Keycloak | Authentication (OAuth2/PKCE, SSO) | `http://localhost:8080` |
 | Semgrep | SAST scanning | CLI tool |
 | Syft | SBOM generation | CLI tool |
-| Chromium | Headless browser (pentesting, PDF) | Managed via Docker |

 ---

@@ -7,23 +7,9 @@ edition = "2021"
 workspace = true

 [dependencies]
-compliance-core = { workspace = true, features = ["mongodb", "telemetry", "axum"] }
-control-map = { workspace = true }
+compliance-core = { workspace = true, features = ["mongodb", "telemetry"] }
 compliance-graph = { path = "../compliance-graph" }
 compliance-dast = { path = "../compliance-dast" }
-# Shared dynamic-execution logic (soft-PLC provisioning + ICS probing), also
-# used by the Werkbank runner.
-werkbank-exec = { path = "../werkbank-exec" }
-# Native firmware build/target detection for bare-metal & RTOS artifacts.
-# Same-company IP, used directly (not via CLI) so the whole tramiton suite is
-# available to the onboarding classifier. NOTE: CI must be able to fetch this
-# private repo (see the git-auth step in .gitea/workflows/ci.yml).
-tramiton-core = { git = "ssh://git@gitea.meghsakha.com:22222/sharang/tramiton.git", tag = "v0.4.1" }
-# tramiton-repro drives the reproducible build (NixBackend seal_and_build) that
-# yields a sealed lock; `libraries_from_inputs` is the analysis-only fallback.
-tramiton-repro = { git = "ssh://git@gitea.meghsakha.com:22222/sharang/tramiton.git", tag = "v0.4.1" }
-# tramiton-sbom renders the bill of materials from a sealed lock (+ binary SCA).
-tramiton-sbom = { git = "ssh://git@gitea.meghsakha.com:22222/sharang/tramiton.git", tag = "v0.4.1" }
 serde = { workspace = true }
 serde_json = { workspace = true }
 tokio = { workspace = true }
@@ -38,16 +24,14 @@ hex = { workspace = true }
 uuid = { workspace = true }
 secrecy = { workspace = true }
 regex = { workspace = true }
-axum = { version = "0.8", features = ["multipart"] }
-tower-http = { version = "0.6", features = ["cors", "trace", "set-header"] }
+axum = "0.8"
+tower-http = { version = "0.6", features = ["cors", "trace"] }
 git2 = "0.20"
 octocrab = "0.44"
 tokio-cron-scheduler = "0.13"
 dotenvy = "0.15"
 hmac = "0.12"
 walkdir = "2"
-# Read-only XML tree parsing for PLCopen project files (POU extraction).
-roxmltree = "0.20"
 base64 = "0.22"
 urlencoding = "2"
 futures-util = "0.3"
@@ -58,16 +42,3 @@ tokio-tungstenite = { version = "0.26", features = ["rustls-tls-webpki-roots"] }
 futures-core = "0.3"
 dashmap = { workspace = true }
 tokio-stream = { workspace = true }
-rand = { workspace = true }
-
-[dev-dependencies]
-compliance-core = { workspace = true, features = ["mongodb", "axum"] }
-tower = { version = "0.5", features = ["util"] }
-reqwest = { workspace = true }
-serde_json = { workspace = true }
-tokio = { workspace = true }
-mongodb = { workspace = true }
-uuid = { workspace = true }
-secrecy = { workspace = true }
-axum = { version = "0.8", features = ["multipart"] }
-tower-http = { version = "0.6", features = ["cors"] }
@@ -1,117 +0,0 @@
-# Custom semgrep rules for CRA controls that no off-the-shelf ruleset digs out.
-# Each rule id is `cra-ai-<n>-<slug>` and is keyed back to its control via the
-# `control-map` LUT (by rule-id suffix, so semgrep's path prefix on check_id does
-# not matter). Detection here is deterministic; the grounded LLM judge downstream
-# only confirms/refutes — it never detects. Keep patterns tight: a false positive
-# that the judge refutes marks the whole finding a false positive.
-rules:
-  # --- cra-ai-1: Secure-by-Default-Konfiguration -------------------------------
-  - id: cra-ai-1-flask-debug-enabled
-    languages: [python]
-    severity: WARNING
-    message: Flask app started with debug=True — ships an interactive debugger / code execution in production (secure-by-default violation).
-    metadata:
-      cwe: ["CWE-489: Active Debug Code"]
-      control: cra-ai-1
-    patterns:
-      - pattern: '$APP.run(..., debug=True, ...)'
-
-  - id: cra-ai-1-django-debug-true
-    languages: [python]
-    severity: WARNING
-    message: Django DEBUG = True — leaks stack traces / settings in production (secure-by-default violation).
-    metadata:
-      cwe: ["CWE-489: Active Debug Code"]
-      control: cra-ai-1
-    patterns:
-      - pattern: 'DEBUG = True'
-
-  - id: cra-ai-1-tls-verify-disabled
-    languages: [python]
-    severity: ERROR
-    message: TLS certificate verification disabled (verify=False) — defeats transport security by default.
-    metadata:
-      cwe: ["CWE-295: Improper Certificate Validation"]
-      control: cra-ai-1
-    patterns:
-      - pattern: 'requests.$M(..., verify=False, ...)'
-
-  - id: cra-ai-1-cors-wildcard
-    languages: [javascript, typescript]
-    severity: WARNING
-    message: CORS Access-Control-Allow-Origin set to "*" — opens the API to any origin by default.
-    metadata:
-      cwe: ["CWE-942: Permissive Cross-domain Policy with Untrusted Domains"]
-      control: cra-ai-1
-    patterns:
-      - pattern-either:
-          - pattern: '$RES.header("Access-Control-Allow-Origin", "*")'
-          - pattern: '$RES.setHeader("Access-Control-Allow-Origin", "*")'
-
-  # --- cra-ai-7: Starke Authentifizierung (weak password hashing) --------------
-  - id: cra-ai-7-weak-password-hash
-    languages: [python]
-    severity: ERROR
-    message: Password/secret hashed with a fast, broken digest (md5/sha1) — use a password KDF (bcrypt/scrypt/argon2).
-    metadata:
-      cwe: ["CWE-916: Use of Password Hash With Insufficient Computational Effort"]
-      control: cra-ai-7
-    patterns:
-      - pattern-either:
-          - pattern: 'hashlib.md5($PW)'
-          - pattern: 'hashlib.sha1($PW)'
-      - metavariable-regex:
-          metavariable: $PW
-          regex: '(?i).*(pass|pwd|secret|cred|token).*'
-
-  # --- cra-ai-10: Sitzungsmanagement (insecure session cookies) ----------------
-  - id: cra-ai-10-session-cookie-insecure
-    languages: [python]
-    severity: ERROR
-    message: Session cookie hardened flag explicitly disabled (Secure/HttpOnly = False) — session token exposed to theft.
-    metadata:
-      cwe: ["CWE-614: Sensitive Cookie in HTTPS Session Without 'Secure' Attribute"]
-      control: cra-ai-10
-    patterns:
-      - pattern-either:
-          - pattern: 'SESSION_COOKIE_SECURE = False'
-          - pattern: 'SESSION_COOKIE_HTTPONLY = False'
-
-  - id: cra-ai-10-express-cookie-insecure
-    languages: [javascript, typescript]
-    severity: ERROR
-    message: Express cookie set with secure/httpOnly = false — session token exposed to interception / XSS theft.
-    metadata:
-      cwe: ["CWE-614: Sensitive Cookie in HTTPS Session Without 'Secure' Attribute"]
-      control: cra-ai-10
-    patterns:
-      - pattern-either:
-          - pattern: '$RES.cookie($NAME, $VAL, {..., secure: false, ...})'
-          - pattern: '$RES.cookie($NAME, $VAL, {..., httpOnly: false, ...})'
-
-  # --- cra-ai-14: Speicher-Schutz / Data at Rest (weak cipher) -----------------
-  - id: cra-ai-14-python-weak-cipher
-    languages: [python]
-    severity: ERROR
-    message: Data-at-rest encrypted with a broken cipher/mode (ECB, DES, 3DES) — provides no real confidentiality.
-    metadata:
-      cwe: ["CWE-327: Use of a Broken or Risky Cryptographic Algorithm"]
-      control: cra-ai-14
-    patterns:
-      - pattern-either:
-          - pattern: 'AES.new($K, AES.MODE_ECB, ...)'
-          - pattern: 'DES.new(...)'
-          - pattern: 'DES3.new(...)'
-
-  - id: cra-ai-14-node-weak-cipher
-    languages: [javascript, typescript]
-    severity: ERROR
-    message: Data-at-rest encrypted with a broken cipher (DES / deprecated createCipher) — provides no real confidentiality.
-    metadata:
-      cwe: ["CWE-327: Use of a Broken or Risky Cryptographic Algorithm"]
-      control: cra-ai-14
-    patterns:
-      - pattern-either:
-          - pattern: 'crypto.createCipheriv("des-ecb", ...)'
-          - pattern: 'crypto.createCipheriv("des", ...)'
-          - pattern: 'crypto.createCipher(...)'
@@ -6,7 +6,7 @@ use tokio::sync::{broadcast, watch, Semaphore};
 use compliance_core::models::pentest::PentestEvent;
 use compliance_core::AgentConfig;

-use crate::database::DatabasePool;
+use crate::database::Database;
 use crate::llm::LlmClient;
 use crate::pipeline::orchestrator::PipelineOrchestrator;

@@ -16,10 +16,7 @@ const DEFAULT_MAX_CONCURRENT_SESSIONS: usize = 5;
 #[derive(Clone)]
 pub struct ComplianceAgent {
    pub config: AgentConfig,
-    /// Per-tenant Mongo broker. Every code path must obtain a
-    /// tenant-scoped [`crate::database::Database`] from this pool —
-    /// there is no single shared database any more.
-    pub db_pool: DatabasePool,
+    pub db: Database,
    pub llm: Arc<LlmClient>,
    pub http: reqwest::Client,
    /// Per-session broadcast senders for SSE streaming.
@@ -31,23 +28,18 @@ pub struct ComplianceAgent {
 }

 impl ComplianceAgent {
-    pub fn new(config: AgentConfig, db_pool: DatabasePool) -> Self {
+    pub fn new(config: AgentConfig, db: Database) -> Self {
        let llm = Arc::new(LlmClient::new(
            config.litellm_url.clone(),
            config.litellm_api_key.clone(),
            config.litellm_model.clone(),
            config.litellm_embed_model.clone(),
        ));
-        let http = reqwest::Client::builder()
-            .timeout(std::time::Duration::from_secs(30))
-            .connect_timeout(std::time::Duration::from_secs(10))
-            .build()
-            .unwrap_or_default();
        Self {
            config,
-            db_pool,
+            db,
            llm,
-            http,
+            http: reqwest::Client::new(),
            session_streams: Arc::new(DashMap::new()),
            session_pause: Arc::new(DashMap::new()),
            session_semaphore: Arc::new(Semaphore::new(DEFAULT_MAX_CONCURRENT_SESSIONS)),
@@ -56,57 +48,44 @@ impl ComplianceAgent {

    pub async fn run_scan(
        &self,
-        tenant_id: &str,
        repo_id: &str,
        trigger: compliance_core::models::ScanTrigger,
    ) -> Result<(), crate::error::AgentError> {
-        let db = self.db_pool.for_tenant_id(tenant_id).await?;
-        let orchestrator =
-            PipelineOrchestrator::new(self.config.clone(), db, self.llm.clone(), self.http.clone());
-        orchestrator.run_target(repo_id, trigger).await
-    }
-
-    /// Alias for [`Self::run_scan`] — every scan runs the unified onboarded-target
-    /// pipeline. Kept as a distinct name for the `/targets/{id}/scan` endpoint's
-    /// intent.
-    pub async fn run_target_scan(
-        &self,
-        tenant_id: &str,
-        target_id: &str,
-        trigger: compliance_core::models::ScanTrigger,
-    ) -> Result<(), crate::error::AgentError> {
-        let db = self.db_pool.for_tenant_id(tenant_id).await?;
-        let orchestrator =
-            PipelineOrchestrator::new(self.config.clone(), db, self.llm.clone(), self.http.clone());
-        orchestrator.run_target(target_id, trigger).await
+        let orchestrator = PipelineOrchestrator::new(
+            self.config.clone(),
+            self.db.clone(),
+            self.llm.clone(),
+            self.http.clone(),
+        );
+        orchestrator.run(repo_id, trigger).await
    }

    /// Run a PR review: scan the diff and post review comments.
    pub async fn run_pr_review(
        &self,
-        tenant_id: &str,
        repo_id: &str,
        pr_number: u64,
        base_sha: &str,
        head_sha: &str,
    ) -> Result<(), crate::error::AgentError> {
-        let db = self.db_pool.for_tenant_id(tenant_id).await?;
-        let oid = mongodb::bson::oid::ObjectId::parse_str(repo_id)
-            .map_err(|e| crate::error::AgentError::Other(e.to_string()))?;
-        let target = db
-            .onboarded_targets()
-            .find_one(mongodb::bson::doc! { "_id": oid })
+        let repo = self
+            .db
+            .repositories()
+            .find_one(mongodb::bson::doc! {
+                "_id": mongodb::bson::oid::ObjectId::parse_str(repo_id)
+                    .map_err(|e| crate::error::AgentError::Other(e.to_string()))?
+            })
            .await?
            .ok_or_else(|| {
-                crate::error::AgentError::Other(format!("Target {repo_id} not found"))
+                crate::error::AgentError::Other(format!("Repository {repo_id} not found"))
            })?;
-        let code = target.code_artifact().ok_or_else(|| {
-            crate::error::AgentError::Other(format!("Target {repo_id} has no code artifact"))
-        })?;
-        let repo = crate::pipeline::repo_view::RepoView::from_target(&target, code);

-        let orchestrator =
-            PipelineOrchestrator::new(self.config.clone(), db, self.llm.clone(), self.http.clone());
+        let orchestrator = PipelineOrchestrator::new(
+            self.config.clone(),
+            self.db.clone(),
+            self.llm.clone(),
+            self.http.clone(),
+        );
        orchestrator
            .run_pr_review(&repo, repo_id, pr_number, base_sha, head_sha)
            .await
@@ -0,0 +1,113 @@
+use std::sync::Arc;
+
+use axum::{
+    extract::Request,
+    middleware::Next,
+    response::{IntoResponse, Response},
+};
+use jsonwebtoken::{decode, decode_header, jwk::JwkSet, DecodingKey, Validation};
+use reqwest::StatusCode;
+use serde::Deserialize;
+use tokio::sync::RwLock;
+
+/// Cached JWKS from Keycloak for token validation.
+#[derive(Clone)]
+pub struct JwksState {
+    pub jwks: Arc<RwLock<Option<JwkSet>>>,
+    pub jwks_url: String,
+}
+
+#[derive(Debug, Deserialize)]
+struct Claims {
+    #[allow(dead_code)]
+    sub: String,
+}
+
+const PUBLIC_ENDPOINTS: &[&str] = &["/api/v1/health"];
+
+/// Middleware that validates Bearer JWT tokens against Keycloak's JWKS.
+///
+/// Skips validation for health check endpoints.
+/// If `JwksState` is not present as an extension (keycloak not configured),
+/// all requests pass through.
+pub async fn require_jwt_auth(request: Request, next: Next) -> Response {
+    let path = request.uri().path();
+
+    if PUBLIC_ENDPOINTS.contains(&path) {
+        return next.run(request).await;
+    }
+
+    let jwks_state = match request.extensions().get::<JwksState>() {
+        Some(s) => s.clone(),
+        None => return next.run(request).await,
+    };
+
+    let auth_header = match request.headers().get("authorization") {
+        Some(h) => h,
+        None => return (StatusCode::UNAUTHORIZED, "Missing authorization header").into_response(),
+    };
+
+    let token = match auth_header.to_str() {
+        Ok(s) if s.starts_with("Bearer ") => &s[7..],
+        _ => return (StatusCode::UNAUTHORIZED, "Invalid authorization header").into_response(),
+    };
+
+    match validate_token(token, &jwks_state).await {
+        Ok(()) => next.run(request).await,
+        Err(e) => {
+            tracing::warn!("JWT validation failed: {e}");
+            (StatusCode::UNAUTHORIZED, "Invalid token").into_response()
+        }
+    }
+}
+
+async fn validate_token(token: &str, state: &JwksState) -> Result<(), String> {
+    let header = decode_header(token).map_err(|e| format!("failed to decode JWT header: {e}"))?;
+
+    let kid = header
+        .kid
+        .ok_or_else(|| "JWT missing kid header".to_string())?;
+
+    let jwks = fetch_or_get_jwks(state).await?;
+
+    let jwk = jwks
+        .keys
+        .iter()
+        .find(|k| k.common.key_id.as_deref() == Some(&kid))
+        .ok_or_else(|| "no matching key found in JWKS".to_string())?;
+
+    let decoding_key =
+        DecodingKey::from_jwk(jwk).map_err(|e| format!("failed to create decoding key: {e}"))?;
+
+    let mut validation = Validation::new(header.alg);
+    validation.validate_exp = true;
+    validation.validate_aud = false;
+
+    decode::<Claims>(token, &decoding_key, &validation)
+        .map_err(|e| format!("token validation failed: {e}"))?;
+
+    Ok(())
+}
+
+async fn fetch_or_get_jwks(state: &JwksState) -> Result<JwkSet, String> {
+    {
+        let cached = state.jwks.read().await;
+        if let Some(ref jwks) = *cached {
+            return Ok(jwks.clone());
+        }
+    }
+
+    let resp = reqwest::get(&state.jwks_url)
+        .await
+        .map_err(|e| format!("failed to fetch JWKS: {e}"))?;
+
+    let jwks: JwkSet = resp
+        .json()
+        .await
+        .map_err(|e| format!("failed to parse JWKS: {e}"))?;
+
+    let mut cached = state.jwks.write().await;
+    *cached = Some(jwks.clone());
+
+    Ok(jwks)
+}
@@ -1,115 +0,0 @@
-//! Cross-tenant admin endpoints (`/api/v1/admin/*`).
-//!
-//! Operator-only. Auth is a **static bearer token** (`ADMIN_API_TOKEN`
-//! env on the agent) — explicitly NOT a Keycloak JWT, because the
-//! whole point of these endpoints is to operate ACROSS tenants. A
-//! customer JWT (which always carries a single tenant_id) has no
-//! business mounting them.
-//!
-//! Routes are only registered when `ADMIN_API_TOKEN` is set. With no
-//! token, the endpoints don't exist at all (404), which is a stronger
-//! guarantee than "401 if you guess the path".
-//!
-//! Operations:
-//! - `GET    /api/v1/admin/tenants`              — list tenant DBs
-//! - `DELETE /api/v1/admin/tenants/{tenant_id}`  — GDPR delete
-//!
-//! Tenant ids in URLs are passed as-is to `DatabasePool::drop_tenant`,
-//! which sanitises them the same way it does for creation. Listing
-//! returns the raw DB names from `list_tenant_db_names` — operators
-//! can reverse-derive the tenant_id from the prefix.
-
-use axum::extract::{Extension, Path, Request};
-use axum::http::{header, StatusCode};
-use axum::middleware::Next;
-use axum::response::{IntoResponse, Response};
-use axum::Json;
-use secrecy::ExposeSecret;
-use serde::Serialize;
-
-use super::dto::AgentExt;
-
-#[derive(Serialize)]
-pub struct ListTenantDbsResponse {
-    pub tenant_db_names: Vec<String>,
-}
-
-#[tracing::instrument(skip_all)]
-pub async fn list_tenant_dbs(
-    Extension(agent): AgentExt,
-) -> Result<Json<ListTenantDbsResponse>, StatusCode> {
-    let names = agent.db_pool.list_tenant_db_names().await.map_err(|e| {
-        tracing::error!("admin: list_tenant_db_names failed: {e}");
-        StatusCode::INTERNAL_SERVER_ERROR
-    })?;
-    Ok(Json(ListTenantDbsResponse {
-        tenant_db_names: names,
-    }))
-}
-
-#[tracing::instrument(skip_all, fields(tenant_id = %tenant_id))]
-pub async fn drop_tenant_db(
-    Extension(agent): AgentExt,
-    Path(tenant_id): Path<String>,
-) -> Result<Json<serde_json::Value>, StatusCode> {
-    agent.db_pool.drop_tenant(&tenant_id).await.map_err(|e| {
-        tracing::error!("admin: drop_tenant failed: {e}");
-        StatusCode::INTERNAL_SERVER_ERROR
-    })?;
-    Ok(Json(serde_json::json!({ "status": "dropped" })))
-}
-
-/// Constant-time-ish comparison of the configured admin token against
-/// the incoming bearer. Uses `subtle`-style byte equality so timing
-/// attacks can't probe the token character by character.
-fn tokens_eq(a: &str, b: &str) -> bool {
-    if a.len() != b.len() {
-        return false;
-    }
-    let mut diff = 0u8;
-    for (x, y) in a.bytes().zip(b.bytes()) {
-        diff |= x ^ y;
-    }
-    diff == 0
-}
-
-/// Middleware enforcing the static `ADMIN_API_TOKEN`. Mounted only on
-/// the admin sub-router, so this never runs on customer routes.
-pub async fn require_admin_token(
-    Extension(agent): AgentExt,
-    request: Request,
-    next: Next,
-) -> Response {
-    let Some(expected) = agent.config.admin_api_token.as_ref() else {
-        // Belt-and-braces — if the routes were somehow mounted without
-        // a token configured, refuse rather than no-op-pass.
-        return (StatusCode::NOT_FOUND, "admin disabled").into_response();
-    };
-    let presented = request
-        .headers()
-        .get(header::AUTHORIZATION)
-        .and_then(|v| v.to_str().ok())
-        .and_then(|s| s.strip_prefix("Bearer "))
-        .map(|s| s.trim());
-    let Some(presented) = presented.filter(|s| !s.is_empty()) else {
-        return (StatusCode::UNAUTHORIZED, "Missing bearer token").into_response();
-    };
-    if !tokens_eq(presented, expected.expose_secret()) {
-        return (StatusCode::UNAUTHORIZED, "Invalid admin token").into_response();
-    }
-    next.run(request).await
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn tokens_eq_basic() {
-        assert!(tokens_eq("abc", "abc"));
-        assert!(!tokens_eq("abc", "abd"));
-        assert!(!tokens_eq("abc", "abcd"));
-        assert!(!tokens_eq("", "x"));
-        assert!(tokens_eq("", ""));
-    }
-}
@@ -7,13 +7,11 @@ use mongodb::bson::doc;

 use compliance_core::models::chat::{ChatRequest, ChatResponse, SourceReference};
 use compliance_core::models::embedding::EmbeddingBuildRun;
-use compliance_core::tenant_ctx::TenantCtx;
 use compliance_graph::graph::embedding_store::EmbeddingStore;

 use crate::agent::ComplianceAgent;
 use crate::rag::pipeline::RagPipeline;

-use super::dto::tenant_db;
 use super::ApiResponse;

 type AgentExt = Extension<Arc<ComplianceAgent>>;
@@ -22,12 +20,10 @@ type AgentExt = Extension<Arc<ComplianceAgent>>;
 #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
 pub async fn chat(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(repo_id): Path<String>,
    Json(req): Json<ChatRequest>,
 ) -> Result<Json<ApiResponse<ChatResponse>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let pipeline = RagPipeline::new(agent.llm.clone(), db.inner());
+    let pipeline = RagPipeline::new(agent.llm.clone(), agent.db.inner());

    // Step 1: Embed the user's message
    let query_vectors = agent
@@ -94,13 +90,10 @@ pub async fn chat(
    };

    let system_prompt = format!(
-        "You are a code assistant for this repository. Answer questions using the code context below.\n\n\
-         Rules:\n\
-         - Reference specific files, functions, and line numbers\n\
-         - Show code snippets when they help explain the answer\n\
-         - If the context is insufficient, say what's missing rather than guessing\n\
-         - Be concise — lead with the answer, then explain if needed\n\
-         - For security questions, note relevant CWEs and link to the finding if one exists\n\n\
+        "You are an expert code assistant for a software repository. \
+         Answer the user's question based on the code context below. \
+         Reference specific files and functions when relevant. \
+         If the context doesn't contain enough information, say so.\n\n\
         ## Code Context\n\n{code_context}"
    );

@@ -137,16 +130,13 @@ pub async fn chat(
 #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
 pub async fn build_embeddings(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(repo_id): Path<String>,
 ) -> Result<Json<serde_json::Value>, StatusCode> {
-    // Resolve the tenant DB up front so we can move it into the spawn;
-    // the JWT/dev context isn't available inside detached tasks.
-    let db = tenant_db(&agent, &tenant).await?;
    let agent_clone = (*agent).clone();
    tokio::spawn(async move {
-        let repo = match db
-            .onboarded_targets()
+        let repo = match agent_clone
+            .db
+            .repositories()
            .find_one(doc! { "_id": mongodb::bson::oid::ObjectId::parse_str(&repo_id).ok() })
            .await
        {
@@ -158,7 +148,8 @@ pub async fn build_embeddings(
        };

        // Get latest graph build
-        let build = match db
+        let build = match agent_clone
+            .db
            .graph_builds()
            .find_one(doc! { "repo_id": &repo_id })
            .sort(doc! { "started_at": -1 })
@@ -177,39 +168,35 @@ pub async fn build_embeddings(
            .unwrap_or_else(|| "unknown".to_string());

        // Get nodes
-        let nodes: Vec<compliance_core::models::graph::CodeNode> =
-            match db.graph_nodes().find(doc! { "repo_id": &repo_id }).await {
-                Ok(cursor) => {
-                    use futures_util::StreamExt;
-                    let mut items = Vec::new();
-                    let mut cursor = cursor;
-                    while let Some(Ok(item)) = cursor.next().await {
-                        items.push(item);
-                    }
-                    items
+        let nodes: Vec<compliance_core::models::graph::CodeNode> = match agent_clone
+            .db
+            .graph_nodes()
+            .find(doc! { "repo_id": &repo_id })
+            .await
+        {
+            Ok(cursor) => {
+                use futures_util::StreamExt;
+                let mut items = Vec::new();
+                let mut cursor = cursor;
+                while let Some(Ok(item)) = cursor.next().await {
+                    items.push(item);
                }
-                Err(e) => {
-                    tracing::error!("[{repo_id}] Failed to fetch nodes: {e}");
-                    return;
-                }
-            };
-
-        let code = match repo.code_artifact() {
-            Some(c) => c,
-            None => {
-                tracing::error!("Target {repo_id} has no code artifact for embedding build");
+                items
+            }
+            Err(e) => {
+                tracing::error!("[{repo_id}] Failed to fetch nodes: {e}");
                return;
            }
        };
-        let view = crate::pipeline::repo_view::RepoView::from_target(&repo, code);
+
        let creds = crate::pipeline::git::RepoCredentials {
            ssh_key_path: Some(agent_clone.config.ssh_key_path.clone()),
-            auth_token: view.auth_token.clone(),
-            auth_username: view.auth_username.clone(),
+            auth_token: repo.auth_token.clone(),
+            auth_username: repo.auth_username.clone(),
        };
        let git_ops =
            crate::pipeline::git::GitOps::new(&agent_clone.config.git_clone_base_path, creds);
-        let repo_path = match git_ops.clone_or_fetch(&view.git_url, &view.name) {
+        let repo_path = match git_ops.clone_or_fetch(&repo.git_url, &repo.name) {
            Ok(p) => p,
            Err(e) => {
                tracing::error!("Failed to clone repo for embedding build: {e}");
@@ -217,7 +204,7 @@ pub async fn build_embeddings(
            }
        };

-        let pipeline = RagPipeline::new(agent_clone.llm.clone(), db.inner());
+        let pipeline = RagPipeline::new(agent_clone.llm.clone(), agent_clone.db.inner());
        match pipeline
            .build_embeddings(&repo_id, &repo_path, &graph_build_id, &nodes)
            .await
@@ -244,11 +231,9 @@ pub async fn build_embeddings(
 #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
 pub async fn embedding_status(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(repo_id): Path<String>,
 ) -> Result<Json<ApiResponse<Option<EmbeddingBuildRun>>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let store = EmbeddingStore::new(db.inner());
+    let store = EmbeddingStore::new(agent.db.inner());
    let build = store.get_latest_build(&repo_id).await.map_err(|e| {
        tracing::error!("Failed to get embedding status: {e}");
        StatusCode::INTERNAL_SERVER_ERROR
@@ -7,11 +7,9 @@ use mongodb::bson::doc;
 use serde::Deserialize;

 use compliance_core::models::dast::{DastFinding, DastScanRun, DastTarget, DastTargetType};
-use compliance_core::tenant_ctx::TenantCtx;

 use crate::agent::ComplianceAgent;

-use super::dto::tenant_db;
 use super::{collect_cursor_async, ApiResponse, PaginationParams};

 type AgentExt = Extension<Arc<ComplianceAgent>>;
@@ -47,11 +45,9 @@ fn default_rate_limit() -> u32 {
 #[tracing::instrument(skip_all)]
 pub async fn list_targets(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(params): Query<PaginationParams>,
 ) -> Result<Json<ApiResponse<Vec<DastTarget>>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let skip = (params.page.saturating_sub(1)) * params.limit as u64;
    let total = db
        .dast_targets()
@@ -84,7 +80,6 @@ pub async fn list_targets(
 #[tracing::instrument(skip_all)]
 pub async fn add_target(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Json(req): Json<AddTargetRequest>,
 ) -> Result<Json<ApiResponse<DastTarget>>, StatusCode> {
    let mut target = DastTarget::new(req.name, req.base_url, req.target_type);
@@ -94,8 +89,9 @@ pub async fn add_target(
    target.rate_limit = req.rate_limit;
    target.allow_destructive = req.allow_destructive;

-    let db = tenant_db(&agent, &tenant).await?;
-    db.dast_targets()
+    agent
+        .db
+        .dast_targets()
        .insert_one(&target)
        .await
        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
@@ -111,19 +107,19 @@ pub async fn add_target(
 #[tracing::instrument(skip_all, fields(target_id = %id))]
 pub async fn trigger_scan(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
 ) -> Result<Json<serde_json::Value>, StatusCode> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;

-    let target = db
+    let target = agent
+        .db
        .dast_targets()
        .find_one(doc! { "_id": oid })
        .await
        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?
        .ok_or(StatusCode::NOT_FOUND)?;

+    let db = agent.db.clone();
    tokio::spawn(async move {
        let orchestrator = compliance_dast::DastOrchestrator::new(100);
        match orchestrator.run_scan(&target, Vec::new()).await {
@@ -151,11 +147,9 @@ pub async fn trigger_scan(
 #[tracing::instrument(skip_all)]
 pub async fn list_scan_runs(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(params): Query<PaginationParams>,
 ) -> Result<Json<ApiResponse<Vec<DastScanRun>>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let skip = (params.page.saturating_sub(1)) * params.limit as u64;
    let total = db
        .dast_scan_runs()
@@ -189,11 +183,9 @@ pub async fn list_scan_runs(
 #[tracing::instrument(skip_all)]
 pub async fn list_findings(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(params): Query<PaginationParams>,
 ) -> Result<Json<ApiResponse<Vec<DastFinding>>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let skip = (params.page.saturating_sub(1)) * params.limit as u64;
    let total = db
        .dast_findings()
@@ -227,13 +219,12 @@ pub async fn list_findings(
 #[tracing::instrument(skip_all, fields(finding_id = %id))]
 pub async fn get_finding(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
 ) -> Result<Json<ApiResponse<DastFinding>>, StatusCode> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;

-    let finding = db
+    let finding = agent
+        .db
        .dast_findings()
        .find_one(doc! { "_id": oid })
        .await
@@ -180,27 +180,6 @@ pub struct SbomVersionDiff {
 pub(crate) type AgentExt = axum::extract::Extension<std::sync::Arc<crate::agent::ComplianceAgent>>;
 pub(crate) type ApiResult<T> = Result<axum::Json<ApiResponse<T>>, axum::http::StatusCode>;

-/// Resolve a tenant-scoped [`Database`] from the request's
-/// [`TenantContext`] (inserted by the M7.1 JWT middleware, or by the
-/// dev fallback in unsecured environments). The pool ensures the
-/// tenant's indexes idempotently.
-///
-/// Returns 500 on the rare path where Mongo refuses the database
-/// handle — the M7.1 auth/status middleware already rejects every
-/// other failure mode with 4xx before we get here.
-pub(crate) async fn tenant_db(
-    agent: &crate::agent::ComplianceAgent,
-    tenant: &compliance_core::tenant_ctx::TenantCtx,
-) -> Result<crate::database::Database, axum::http::StatusCode> {
-    agent.db_pool.for_tenant(&tenant.0).await.map_err(|e| {
-        tracing::error!(
-            tenant_id = %tenant.0.tenant_id,
-            "Failed to acquire tenant database: {e}"
-        );
-        axum::http::StatusCode::INTERNAL_SERVER_ERROR
-    })
-}
-
 pub(crate) async fn collect_cursor_async<T: serde::de::DeserializeOwned + Unpin + Send>(
    mut cursor: mongodb::Cursor<T>,
 ) -> Vec<T> {
@@ -5,16 +5,13 @@ use mongodb::bson::doc;

 use super::dto::*;
 use compliance_core::models::Finding;
-use compliance_core::tenant_ctx::TenantCtx;

 #[tracing::instrument(skip_all, fields(repo_id = ?filter.repo_id, severity = ?filter.severity, scan_type = ?filter.scan_type))]
 pub async fn list_findings(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(filter): Query<FindingsFilter>,
 ) -> ApiResult<Vec<Finding>> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let mut query = doc! {};
    if let Some(repo_id) = &filter.repo_id {
        query.insert("repo_id", repo_id);
@@ -84,12 +81,11 @@ pub async fn list_findings(
 #[tracing::instrument(skip_all, fields(finding_id = %id))]
 pub async fn get_finding(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
 ) -> Result<Json<ApiResponse<Finding>>, StatusCode> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;
-    let finding = db
+    let finding = agent
+        .db
        .findings()
        .find_one(doc! { "_id": oid })
        .await
@@ -106,14 +102,14 @@ pub async fn get_finding(
 #[tracing::instrument(skip_all, fields(finding_id = %id))]
 pub async fn update_finding_status(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
    Json(req): Json<UpdateStatusRequest>,
 ) -> Result<Json<serde_json::Value>, StatusCode> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;

-    db.findings()
+    agent
+        .db
+        .findings()
        .update_one(
            doc! { "_id": oid },
            doc! { "$set": { "status": &req.status, "updated_at": mongodb::bson::DateTime::now() } },
@@ -127,7 +123,6 @@ pub async fn update_finding_status(
 #[tracing::instrument(skip_all)]
 pub async fn bulk_update_finding_status(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Json(req): Json<BulkUpdateStatusRequest>,
 ) -> Result<Json<serde_json::Value>, StatusCode> {
    let oids: Vec<mongodb::bson::oid::ObjectId> = req
@@ -140,8 +135,8 @@ pub async fn bulk_update_finding_status(
        return Err(StatusCode::BAD_REQUEST);
    }

-    let db = tenant_db(&agent, &tenant).await?;
-    let result = db
+    let result = agent
+        .db
        .findings()
        .update_many(
            doc! { "_id": { "$in": oids } },
@@ -158,14 +153,14 @@ pub async fn bulk_update_finding_status(
 #[tracing::instrument(skip_all)]
 pub async fn update_finding_feedback(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
    Json(req): Json<UpdateFeedbackRequest>,
 ) -> Result<Json<serde_json::Value>, StatusCode> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;

-    db.findings()
+    agent
+        .db
+        .findings()
        .update_one(
            doc! { "_id": oid },
            doc! { "$set": { "developer_feedback": &req.feedback, "updated_at": mongodb::bson::DateTime::now() } },
@@ -7,11 +7,9 @@ use mongodb::bson::doc;
 use serde::{Deserialize, Serialize};

 use compliance_core::models::graph::{CodeEdge, CodeNode, GraphBuildRun, ImpactAnalysis};
-use compliance_core::tenant_ctx::TenantCtx;

 use crate::agent::ComplianceAgent;

-use super::dto::tenant_db;
 use super::{collect_cursor_async, ApiResponse};

 type AgentExt = Extension<Arc<ComplianceAgent>>;
@@ -38,11 +36,9 @@ fn default_search_limit() -> usize {
 #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
 pub async fn get_graph(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(repo_id): Path<String>,
 ) -> Result<Json<ApiResponse<GraphData>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;

    // Get latest build
    let build: Option<GraphBuildRun> = db
@@ -102,11 +98,9 @@ pub async fn get_graph(
 #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
 pub async fn get_nodes(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(repo_id): Path<String>,
 ) -> Result<Json<ApiResponse<Vec<CodeNode>>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let filter = doc! { "repo_id": &repo_id };

    let nodes: Vec<CodeNode> = match db.graph_nodes().find(filter).await {
@@ -129,11 +123,9 @@ pub async fn get_nodes(
 #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
 pub async fn get_communities(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(repo_id): Path<String>,
 ) -> Result<Json<ApiResponse<Vec<CommunityInfo>>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let filter = doc! { "repo_id": &repo_id };

    let nodes: Vec<CodeNode> = match db.graph_nodes().find(filter).await {
@@ -184,11 +176,9 @@ pub struct CommunityInfo {
 #[tracing::instrument(skip_all, fields(repo_id = %repo_id, finding_id = %finding_id))]
 pub async fn get_impact(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path((repo_id, finding_id)): Path<(String, String)>,
 ) -> Result<Json<ApiResponse<Option<ImpactAnalysis>>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let filter = doc! { "repo_id": &repo_id, "finding_id": &finding_id };

    let impact = db
@@ -208,12 +198,10 @@ pub async fn get_impact(
 #[tracing::instrument(skip_all, fields(repo_id = %repo_id, query = %params.q))]
 pub async fn search_symbols(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(repo_id): Path<String>,
    Query(params): Query<SearchParams>,
 ) -> Result<Json<ApiResponse<Vec<CodeNode>>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;

    // Simple text search on qualified_name and name fields
    let filter = doc! {
@@ -246,16 +234,14 @@ pub async fn search_symbols(
 #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
 pub async fn get_file_content(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(repo_id): Path<String>,
    Query(params): Query<FileContentParams>,
 ) -> Result<Json<ApiResponse<FileContent>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;

    // Look up the repository to get repo name
    let repo = db
-        .onboarded_targets()
+        .repositories()
        .find_one(doc! { "_id": mongodb::bson::oid::ObjectId::parse_str(&repo_id).ok() })
        .await
        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?
@@ -310,14 +296,13 @@ pub struct FileContent {
 #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
 pub async fn trigger_build(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(repo_id): Path<String>,
 ) -> Result<Json<serde_json::Value>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
    let agent_clone = (*agent).clone();
    tokio::spawn(async move {
-        let repo = match db
-            .onboarded_targets()
+        let repo = match agent_clone
+            .db
+            .repositories()
            .find_one(doc! { "_id": mongodb::bson::oid::ObjectId::parse_str(&repo_id).ok() })
            .await
        {
@@ -328,22 +313,14 @@ pub async fn trigger_build(
            }
        };

-        let code = match repo.code_artifact() {
-            Some(c) => c,
-            None => {
-                tracing::error!("Target {repo_id} has no code artifact for graph build");
-                return;
-            }
-        };
-        let view = crate::pipeline::repo_view::RepoView::from_target(&repo, code);
        let creds = crate::pipeline::git::RepoCredentials {
            ssh_key_path: Some(agent_clone.config.ssh_key_path.clone()),
-            auth_token: view.auth_token.clone(),
-            auth_username: view.auth_username.clone(),
+            auth_token: repo.auth_token.clone(),
+            auth_username: repo.auth_username.clone(),
        };
        let git_ops =
            crate::pipeline::git::GitOps::new(&agent_clone.config.git_clone_base_path, creds);
-        let repo_path = match git_ops.clone_or_fetch(&view.git_url, &view.name) {
+        let repo_path = match git_ops.clone_or_fetch(&repo.git_url, &repo.name) {
            Ok(p) => p,
            Err(e) => {
                tracing::error!("Failed to clone repo for graph build: {e}");
@@ -356,7 +333,8 @@ pub async fn trigger_build(

        match engine.build_graph(&repo_path, &repo_id, &graph_build_id) {
            Ok((code_graph, build_run)) => {
-                let store = compliance_graph::graph::persistence::GraphStore::new(db.inner());
+                let store =
+                    compliance_graph::graph::persistence::GraphStore::new(agent_clone.db.inner());
                let _ = store.delete_repo_graph(&repo_id).await;
                let _ = store
                    .store_graph(&build_run, &code_graph.nodes, &code_graph.edges)
@@ -3,35 +3,18 @@ use mongodb::bson::doc;

 use super::dto::*;
 use compliance_core::models::ScanRun;
-use compliance_core::tenant_ctx::TenantCtx;

 #[tracing::instrument(skip_all)]
 pub async fn health() -> Json<serde_json::Value> {
    Json(serde_json::json!({ "status": "ok" }))
 }

-/// GET /api/v1/settings/ssh-public-key — the agent's SSH deploy public key,
-/// for adding as a read-only deploy key on private git targets.
 #[tracing::instrument(skip_all)]
-pub async fn get_ssh_public_key(
-    axum::extract::Extension(agent): AgentExt,
-) -> Result<Json<serde_json::Value>, axum::http::StatusCode> {
-    let public_path = format!("{}.pub", agent.config.ssh_key_path);
-    let public_key =
-        std::fs::read_to_string(&public_path).map_err(|_| axum::http::StatusCode::NOT_FOUND)?;
-    Ok(Json(serde_json::json!({ "public_key": public_key.trim() })))
-}
-
-#[tracing::instrument(skip_all)]
-pub async fn stats_overview(
-    axum::extract::Extension(agent): AgentExt,
-    tenant: TenantCtx,
-) -> ApiResult<OverviewStats> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+pub async fn stats_overview(axum::extract::Extension(agent): AgentExt) -> ApiResult<OverviewStats> {
+    let db = &agent.db;

    let total_repositories = db
-        .onboarded_targets()
+        .repositories()
        .count_documents(doc! {})
        .await
        .unwrap_or(0);
@@ -1,217 +0,0 @@
-use std::path::{Path, PathBuf};
-use std::sync::OnceLock;
-
-use axum::extract::Extension;
-use axum::http::StatusCode;
-use axum::Json;
-use serde::{Deserialize, Serialize};
-use walkdir::WalkDir;
-
-use super::dto::{AgentExt, ApiResponse};
-
-// ── DTOs ─────────────────────────────────────────────────────────────────────
-
-#[derive(Debug, Deserialize)]
-pub struct HelpChatMessage {
-    pub role: String,
-    pub content: String,
-}
-
-#[derive(Debug, Deserialize)]
-pub struct HelpChatRequest {
-    pub message: String,
-    #[serde(default)]
-    pub history: Vec<HelpChatMessage>,
-}
-
-#[derive(Debug, Serialize)]
-pub struct HelpChatResponse {
-    pub message: String,
-}
-
-// ── Doc cache ────────────────────────────────────────────────────────────────
-
-static DOC_CONTEXT: OnceLock<String> = OnceLock::new();
-
-/// Walk upward from `start` until we find a directory containing both
-/// `README.md` and a `docs/` subdirectory.
-fn find_project_root(start: &Path) -> Option<PathBuf> {
-    let mut current = start.to_path_buf();
-    loop {
-        if current.join("README.md").is_file() && current.join("docs").is_dir() {
-            return Some(current);
-        }
-        if !current.pop() {
-            return None;
-        }
-    }
-}
-
-/// Read README.md + all docs/**/*.md (excluding node_modules).
-fn load_docs(root: &Path) -> String {
-    let mut parts: Vec<String> = Vec::new();
-
-    // Root README first
-    if let Ok(content) = std::fs::read_to_string(root.join("README.md")) {
-        parts.push(format!("<!-- file: README.md -->\n{content}"));
-    }
-
-    // docs/**/*.md, skipping node_modules
-    for entry in WalkDir::new(root.join("docs"))
-        .follow_links(false)
-        .into_iter()
-        .filter_entry(|e| {
-            !e.path()
-                .components()
-                .any(|c| c.as_os_str() == "node_modules")
-        })
-        .filter_map(|e| e.ok())
-    {
-        let path = entry.path();
-        if !path.is_file() {
-            continue;
-        }
-        if path
-            .extension()
-            .and_then(|s| s.to_str())
-            .map(|s| !s.eq_ignore_ascii_case("md"))
-            .unwrap_or(true)
-        {
-            continue;
-        }
-
-        let rel = path.strip_prefix(root).unwrap_or(path);
-        if let Ok(content) = std::fs::read_to_string(path) {
-            parts.push(format!("<!-- file: {} -->\n{content}", rel.display()));
-        }
-    }
-
-    if parts.is_empty() {
-        tracing::warn!(
-            "help_chat: no documentation files found under {}",
-            root.display()
-        );
-    } else {
-        tracing::info!(
-            "help_chat: loaded {} documentation file(s) from {}",
-            parts.len(),
-            root.display()
-        );
-    }
-
-    parts.join("\n\n---\n\n")
-}
-
-/// Returns a reference to the cached doc context string, initialised on
-/// first call via `OnceLock`.
-///
-/// Discovery order:
-/// 1. `HELP_DOCS_PATH` env var (explicit override)
-/// 2. Walk up from the binary location
-/// 3. Current working directory
-/// 4. Common Docker paths (/app, /opt/compliance-scanner)
-fn doc_context() -> &'static str {
-    DOC_CONTEXT.get_or_init(|| {
-        // 1. Explicit env var
-        if let Ok(path) = std::env::var("HELP_DOCS_PATH") {
-            let p = PathBuf::from(&path);
-            if p.join("README.md").is_file() || p.join("docs").is_dir() {
-                tracing::info!("help_chat: loading docs from HELP_DOCS_PATH={path}");
-                return load_docs(&p);
-            }
-            tracing::warn!("help_chat: HELP_DOCS_PATH={path} has no README.md or docs/");
-        }
-
-        // 2. Walk up from binary location
-        let start = std::env::current_exe()
-            .ok()
-            .and_then(|p| p.parent().map(Path::to_path_buf))
-            .unwrap_or_else(|| PathBuf::from("."));
-
-        if let Some(root) = find_project_root(&start) {
-            return load_docs(&root);
-        }
-
-        // 3. Current working directory
-        if let Ok(cwd) = std::env::current_dir() {
-            if let Some(root) = find_project_root(&cwd) {
-                return load_docs(&root);
-            }
-            if cwd.join("README.md").is_file() {
-                return load_docs(&cwd);
-            }
-        }
-
-        // 4. Common Docker/deployment paths
-        for candidate in ["/app", "/opt/compliance-scanner", "/srv/compliance-scanner"] {
-            let p = PathBuf::from(candidate);
-            if p.join("README.md").is_file() || p.join("docs").is_dir() {
-                tracing::info!("help_chat: found docs at {candidate}");
-                return load_docs(&p);
-            }
-        }
-
-        tracing::error!(
-            "help_chat: could not locate project root; doc context will be empty. \
-             Set HELP_DOCS_PATH to the directory containing README.md and docs/"
-        );
-        String::new()
-    })
-}
-
-// ── Handler ──────────────────────────────────────────────────────────────────
-
-/// POST /api/v1/help/chat — Answer questions about the compliance-scanner
-/// using the project documentation as grounding context.
-#[tracing::instrument(skip_all)]
-pub async fn help_chat(
-    Extension(agent): AgentExt,
-    Json(req): Json<HelpChatRequest>,
-) -> Result<Json<ApiResponse<HelpChatResponse>>, StatusCode> {
-    let context = doc_context();
-
-    let system_prompt = if context.is_empty() {
-        "You are a helpful assistant for the Compliance Scanner project. \
-         Answer questions about how to use and configure it. \
-         No documentation was loaded at startup, so rely on your general knowledge."
-            .to_string()
-    } else {
-        format!(
-            "You are a helpful assistant for the Compliance Scanner project. \
-             Answer questions about how to use, configure, and understand it \
-             using the documentation below as your primary source of truth.\n\n\
-             Rules:\n\
-             - Prefer information from the provided docs over general knowledge\n\
-             - Quote or reference the relevant doc section when it helps\n\
-             - If the docs do not cover the topic, say so clearly\n\
-             - Be concise — lead with the answer, then explain if needed\n\
-             - Use markdown formatting for readability\n\n\
-             ## Project Documentation\n\n{context}"
-        )
-    };
-
-    let mut messages: Vec<(String, String)> = Vec::with_capacity(req.history.len() + 2);
-    messages.push(("system".to_string(), system_prompt));
-
-    for msg in &req.history {
-        messages.push((msg.role.clone(), msg.content.clone()));
-    }
-    messages.push(("user".to_string(), req.message));
-
-    let response_text = agent
-        .llm
-        .chat_with_messages(messages, Some(0.3))
-        .await
-        .map_err(|e| {
-            tracing::error!("LLM help chat failed: {e}");
-            StatusCode::INTERNAL_SERVER_ERROR
-        })?;
-
-    Ok(Json(ApiResponse {
-        data: HelpChatResponse {
-            message: response_text,
-        },
-        total: None,
-        page: None,
-    }))
-}
@@ -4,16 +4,13 @@ use mongodb::bson::doc;

 use super::dto::*;
 use compliance_core::models::TrackerIssue;
-use compliance_core::tenant_ctx::TenantCtx;

 #[tracing::instrument(skip_all)]
 pub async fn list_issues(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(params): Query<PaginationParams>,
 ) -> ApiResult<Vec<TrackerIssue>> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let skip = (params.page.saturating_sub(1)) * params.limit as u64;
    let total = db
        .tracker_issues()
@@ -1,186 +0,0 @@
-//! `/api/v1/mcp-tokens` — per-tenant API tokens for the MCP server.
-//!
-//! These are opaque static bearers issued via the dashboard (or a
-//! direct curl with a KC JWT) and copied into LLM clients (Claude
-//! Desktop / Cursor / ChatGPT). The MCP server hashes incoming bearers
-//! and looks them up in the cross-tenant `<prefix>__admin.mcp_tokens`
-//! collection to derive the tenant_id for routing.
-//!
-//! The raw token is shown to the caller exactly once at creation; the
-//! database only ever stores the SHA-256 hash. Revocation is a soft
-//! delete (sets `revoked: true`) so the audit log keeps the record.
-
-use axum::extract::{Extension, Path};
-use axum::http::StatusCode;
-use axum::Json;
-use base64::{engine::general_purpose::URL_SAFE_NO_PAD, Engine as _};
-use compliance_core::models::{McpToken, McpTokenView};
-use compliance_core::tenant_ctx::TenantCtx;
-use mongodb::bson::doc;
-use rand::RngCore;
-use sha2::{Digest, Sha256};
-
-use super::dto::{AgentExt, ApiResponse};
-
-/// Mongo collection name inside the admin DB.
-const COLLECTION: &str = "mcp_tokens";
-
-/// Token prefix the MCP server expects on every bearer.
-const TOKEN_PREFIX: &str = "mcpt_";
-
-/// Bytes of randomness behind each token. 32 → ~256 bits.
-/// Encoded as URL-safe base64 without padding → 43 chars.
-/// Combined with `mcpt_` → 48-char tokens.
-const TOKEN_RAND_BYTES: usize = 32;
-
-#[derive(serde::Deserialize)]
-pub struct CreateMcpTokenRequest {
-    pub name: String,
-}
-
-/// Returned exactly once at creation. The `token` field is gone from
-/// the listing endpoint — the user must save it now.
-#[derive(serde::Serialize)]
-pub struct CreateMcpTokenResponse {
-    pub token: String,
-    pub view: McpTokenView,
-}
-
-/// `POST /api/v1/mcp-tokens` — mint a new token for the caller's tenant.
-#[tracing::instrument(skip_all)]
-pub async fn create_mcp_token(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Json(req): Json<CreateMcpTokenRequest>,
-) -> Result<Json<CreateMcpTokenResponse>, StatusCode> {
-    if req.name.trim().is_empty() {
-        return Err(StatusCode::BAD_REQUEST);
-    }
-    let raw = generate_token();
-    let token_hash = sha256_hex(&raw);
-    let token_prefix: String = raw.chars().take(12).collect();
-
-    let mut token = McpToken {
-        id: None,
-        token_hash,
-        token_prefix,
-        tenant_id: tenant.0.tenant_id.clone(),
-        name: req.name.trim().to_string(),
-        created_by: tenant.0.user_id.clone(),
-        created_at: chrono::Utc::now(),
-        last_used_at: None,
-        revoked: false,
-    };
-
-    let col = agent.db_pool.admin_db().collection::<McpToken>(COLLECTION);
-    let res = col.insert_one(&token).await.map_err(|e| {
-        tracing::error!("Failed to insert MCP token: {e}");
-        StatusCode::INTERNAL_SERVER_ERROR
-    })?;
-    token.id = res.inserted_id.as_object_id();
-
-    Ok(Json(CreateMcpTokenResponse {
-        view: McpTokenView::from(&token),
-        token: raw,
-    }))
-}
-
-/// `GET /api/v1/mcp-tokens` — list tokens for the caller's tenant.
-/// Hash is never returned; only metadata + the 12-char prefix so the
-/// user can identify which row is which.
-#[tracing::instrument(skip_all)]
-pub async fn list_mcp_tokens(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-) -> Result<Json<ApiResponse<Vec<McpTokenView>>>, StatusCode> {
-    let col = agent.db_pool.admin_db().collection::<McpToken>(COLLECTION);
-    let mut cursor = col
-        .find(doc! { "tenant_id": &tenant.0.tenant_id })
-        .sort(doc! { "created_at": -1 })
-        .await
-        .map_err(|e| {
-            tracing::error!("Failed to list MCP tokens: {e}");
-            StatusCode::INTERNAL_SERVER_ERROR
-        })?;
-    let mut out = Vec::new();
-    while cursor.advance().await.map_err(|e| {
-        tracing::warn!("MCP tokens cursor advance failed: {e}");
-        StatusCode::INTERNAL_SERVER_ERROR
-    })? {
-        match cursor.deserialize_current() {
-            Ok(t) => out.push(McpTokenView::from(&t)),
-            Err(e) => tracing::warn!("Failed to deserialize MCP token: {e}"),
-        }
-    }
-    Ok(Json(ApiResponse {
-        data: out,
-        total: None,
-        page: None,
-    }))
-}
-
-/// `DELETE /api/v1/mcp-tokens/{id}` — revoke (soft delete).
-/// Scoped to the caller's tenant: a user can't revoke another tenant's
-/// token even if they guess its id.
-#[tracing::instrument(skip_all, fields(id = %id))]
-pub async fn revoke_mcp_token(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Path(id): Path<String>,
-) -> Result<Json<serde_json::Value>, StatusCode> {
-    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let col = agent.db_pool.admin_db().collection::<McpToken>(COLLECTION);
-    let result = col
-        .update_one(
-            doc! { "_id": oid, "tenant_id": &tenant.0.tenant_id },
-            doc! { "$set": { "revoked": true } },
-        )
-        .await
-        .map_err(|e| {
-            tracing::error!("Failed to revoke MCP token: {e}");
-            StatusCode::INTERNAL_SERVER_ERROR
-        })?;
-    if result.matched_count == 0 {
-        return Err(StatusCode::NOT_FOUND);
-    }
-    Ok(Json(serde_json::json!({ "status": "revoked" })))
-}
-
-/// 32 bytes random → URL-safe base64 → 43 chars, no padding.
-/// Prefixed with `mcpt_` so the MCP server can sniff the format
-/// before bothering with the DB lookup.
-fn generate_token() -> String {
-    let mut bytes = [0u8; TOKEN_RAND_BYTES];
-    rand::rng().fill_bytes(&mut bytes);
-    format!("{TOKEN_PREFIX}{}", URL_SAFE_NO_PAD.encode(bytes))
-}
-
-fn sha256_hex(s: &str) -> String {
-    let mut h = Sha256::new();
-    h.update(s.as_bytes());
-    hex::encode(h.finalize())
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn generated_tokens_are_unique_and_prefixed() {
-        let a = generate_token();
-        let b = generate_token();
-        assert_ne!(a, b);
-        assert!(a.starts_with(TOKEN_PREFIX));
-        assert!(b.starts_with(TOKEN_PREFIX));
-        // 5 + 43 = 48 chars
-        assert_eq!(a.len(), 5 + 43);
-    }
-
-    #[test]
-    fn sha256_is_stable_and_64_hex() {
-        let h = sha256_hex("mcpt_abc");
-        assert_eq!(h.len(), 64);
-        assert!(h.chars().all(|c| c.is_ascii_hexdigit()));
-        assert_eq!(sha256_hex("mcpt_abc"), h);
-    }
-}
@@ -1,26 +1,21 @@
-pub mod admin;
 pub mod chat;
 pub mod dast;
 pub mod dto;
 pub mod findings;
 pub mod graph;
 pub mod health;
-pub mod help_chat;
 pub mod issues;
-pub mod mcp_tokens;
-pub mod notifications;
-pub mod onboarding;
-pub mod oscal;
 pub mod pentest_handlers;
 pub use pentest_handlers as pentest;
+pub mod repos;
 pub mod sbom;
 pub mod scans;
-pub mod werkbank_jobs;

 // Re-export all handler functions so routes.rs can use `handlers::function_name`
 pub use dto::*;
 pub use findings::*;
 pub use health::*;
 pub use issues::*;
+pub use repos::*;
 pub use sbom::*;
 pub use scans::*;
@@ -1,183 +0,0 @@
-use axum::extract::Extension;
-use axum::http::StatusCode;
-use axum::Json;
-use mongodb::bson::doc;
-use serde::Deserialize;
-
-use compliance_core::models::notification::CveNotification;
-use compliance_core::tenant_ctx::TenantCtx;
-
-use super::dto::{tenant_db, AgentExt, ApiResponse};
-
-/// GET /api/v1/notifications — List CVE notifications (newest first)
-#[tracing::instrument(skip_all)]
-pub async fn list_notifications(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    axum::extract::Query(params): axum::extract::Query<NotificationFilter>,
-) -> Result<Json<ApiResponse<Vec<CveNotification>>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let mut filter = doc! {};
-
-    // Filter by status (default: show new + read, exclude dismissed)
-    match params.status.as_deref() {
-        Some("all") => {}
-        Some(s) => {
-            filter.insert("status", s);
-        }
-        None => {
-            filter.insert("status", doc! { "$in": ["new", "read"] });
-        }
-    }
-
-    // Filter by severity
-    if let Some(ref sev) = params.severity {
-        filter.insert("severity", sev.as_str());
-    }
-
-    // Filter by repo
-    if let Some(ref repo_id) = params.repo_id {
-        filter.insert("repo_id", repo_id.as_str());
-    }
-
-    let page = params.page.unwrap_or(1).max(1);
-    let limit = params.limit.unwrap_or(50).min(200);
-    let skip = (page - 1) * limit as u64;
-
-    let total = db
-        .cve_notifications()
-        .count_documents(filter.clone())
-        .await
-        .unwrap_or(0);
-
-    let notifications: Vec<CveNotification> = match db
-        .cve_notifications()
-        .find(filter)
-        .sort(doc! { "created_at": -1 })
-        .skip(skip)
-        .limit(limit)
-        .await
-    {
-        Ok(cursor) => {
-            use futures_util::StreamExt;
-            let mut items = Vec::new();
-            let mut cursor = cursor;
-            while let Some(Ok(n)) = cursor.next().await {
-                items.push(n);
-            }
-            items
-        }
-        Err(e) => {
-            tracing::error!("Failed to list notifications: {e}");
-            return Err(StatusCode::INTERNAL_SERVER_ERROR);
-        }
-    };
-
-    Ok(Json(ApiResponse {
-        data: notifications,
-        total: Some(total),
-        page: Some(page),
-    }))
-}
-
-/// GET /api/v1/notifications/count — Count of unread notifications
-#[tracing::instrument(skip_all)]
-pub async fn notification_count(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-) -> Result<Json<serde_json::Value>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let count = db
-        .cve_notifications()
-        .count_documents(doc! { "status": "new" })
-        .await
-        .unwrap_or(0);
-
-    Ok(Json(serde_json::json!({ "count": count })))
-}
-
-/// PATCH /api/v1/notifications/:id/read — Mark a notification as read
-#[tracing::instrument(skip_all, fields(id = %id))]
-pub async fn mark_read(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    axum::extract::Path(id): axum::extract::Path<String>,
-) -> Result<Json<serde_json::Value>, StatusCode> {
-    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;
-
-    let result = db
-        .cve_notifications()
-        .update_one(
-            doc! { "_id": oid },
-            doc! { "$set": {
-                "status": "read",
-                "read_at": mongodb::bson::DateTime::now(),
-            }},
-        )
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-
-    if result.matched_count == 0 {
-        return Err(StatusCode::NOT_FOUND);
-    }
-    Ok(Json(serde_json::json!({ "status": "read" })))
-}
-
-/// PATCH /api/v1/notifications/:id/dismiss — Dismiss a notification
-#[tracing::instrument(skip_all, fields(id = %id))]
-pub async fn dismiss_notification(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    axum::extract::Path(id): axum::extract::Path<String>,
-) -> Result<Json<serde_json::Value>, StatusCode> {
-    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;
-
-    let result = db
-        .cve_notifications()
-        .update_one(
-            doc! { "_id": oid },
-            doc! { "$set": { "status": "dismissed" } },
-        )
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-
-    if result.matched_count == 0 {
-        return Err(StatusCode::NOT_FOUND);
-    }
-    Ok(Json(serde_json::json!({ "status": "dismissed" })))
-}
-
-/// POST /api/v1/notifications/read-all — Mark all new notifications as read
-#[tracing::instrument(skip_all)]
-pub async fn mark_all_read(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-) -> Result<Json<serde_json::Value>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let result = db
-        .cve_notifications()
-        .update_many(
-            doc! { "status": "new" },
-            doc! { "$set": {
-                "status": "read",
-                "read_at": mongodb::bson::DateTime::now(),
-            }},
-        )
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-
-    Ok(Json(
-        serde_json::json!({ "updated": result.modified_count }),
-    ))
-}
-
-#[derive(Debug, Deserialize)]
-pub struct NotificationFilter {
-    pub status: Option<String>,
-    pub severity: Option<String>,
-    pub repo_id: Option<String>,
-    pub page: Option<u64>,
-    pub limit: Option<i64>,
-}
@@ -1,614 +0,0 @@
-//! Onboarding API — CRUD for unified targets, artifact add, classification, and
-//! the scan-applicability matrix. The wizard (and future integrations) drive
-//! onboarding through these endpoints.
-
-use std::collections::HashMap;
-use std::sync::Arc;
-
-use axum::extract::{Extension, Multipart, Path, Query};
-use axum::http::StatusCode;
-use axum::Json;
-use mongodb::bson::{doc, oid::ObjectId, to_bson};
-use serde::{Deserialize, Serialize};
-
-use compliance_core::models::{
-    Artifact, ArtifactKind, ComplianceProfile, OnboardedTarget, PlcFormat, TargetScanConfig,
-    TargetType,
-};
-use compliance_core::scan_matrix::{applicable_scans, supports_pentest};
-use compliance_core::tenant_ctx::TenantCtx;
-
-use crate::agent::ComplianceAgent;
-use crate::classify::{classify_target, MockFirmwareDetector};
-
-use super::dto::tenant_db;
-use super::{collect_cursor_async, ApiResponse, PaginationParams};
-
-type AgentExt = Extension<Arc<ComplianceAgent>>;
-
-/// A client-supplied artifact spec. The server builds the [`Artifact`] (and its
-/// id) from it, so clients never set internal fields.
-#[derive(Deserialize)]
-pub struct ArtifactInput {
-    pub kind: ArtifactKind,
-    pub source_ref: String,
-    #[serde(default)]
-    pub branch: Option<String>,
-    #[serde(default)]
-    pub plc_format: Option<PlcFormat>,
-}
-
-impl ArtifactInput {
-    fn build(&self) -> Artifact {
-        let s = self.source_ref.clone();
-        match self.kind {
-            ArtifactKind::GitRepo => {
-                Artifact::git_repo(s, self.branch.clone().unwrap_or_else(|| "main".to_string()))
-            }
-            ArtifactKind::LiveUrl => Artifact::live_url(s),
-            ArtifactKind::FirmwareImage => Artifact::firmware_image(s),
-            ArtifactKind::SourceArchive => Artifact::source_archive(s),
-            ArtifactKind::MobilePackage => Artifact::mobile_package(s),
-            ArtifactKind::ContainerImage => Artifact::container_image(s),
-            ArtifactKind::PlcProject => {
-                Artifact::plc_project(s, self.plc_format.unwrap_or(PlcFormat::PlcopenXml))
-            }
-            ArtifactKind::PlaintextDescription => Artifact::plaintext(s),
-        }
-    }
-}
-
-#[derive(Deserialize)]
-pub struct CreateTargetRequest {
-    pub name: String,
-    pub target_type: TargetType,
-    #[serde(default)]
-    pub description: Option<String>,
-    #[serde(default)]
-    pub artifacts: Vec<ArtifactInput>,
-}
-
-#[derive(Deserialize)]
-pub struct UpdateTargetRequest {
-    pub name: Option<String>,
-    pub target_type: Option<TargetType>,
-    pub scan_config: Option<TargetScanConfig>,
-    pub compliance_profile: Option<ComplianceProfile>,
-    pub scan_schedule: Option<String>,
-    /// Replace the target's artifacts wholesale (used by the dashboard editor).
-    #[serde(default)]
-    pub artifacts: Option<Vec<ArtifactInput>>,
-}
-
-/// One applicable-scan option, serialized for the wizard.
-#[derive(Serialize)]
-pub struct ScanOptionDto {
-    pub scan: String,
-    pub default_on: bool,
-    pub rationale: String,
-    pub required_artifact: Option<String>,
-    pub blocked_reason: Option<String>,
-}
-
-#[derive(Serialize)]
-pub struct ApplicableScansResponse {
-    pub scans: Vec<ScanOptionDto>,
-    pub pentest_supported: bool,
-}
-
-fn parse_oid(id: &str) -> Result<ObjectId, StatusCode> {
-    ObjectId::parse_str(id).map_err(|_| StatusCode::BAD_REQUEST)
-}
-
-/// GET /api/v1/targets — list onboarded targets (paginated).
-#[tracing::instrument(skip_all)]
-pub async fn list_targets(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Query(params): Query<PaginationParams>,
-) -> Result<Json<ApiResponse<Vec<OnboardedTarget>>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let skip = (params.page.saturating_sub(1)) * params.limit as u64;
-    let total = db
-        .onboarded_targets()
-        .count_documents(doc! {})
-        .await
-        .unwrap_or(0);
-    let targets = match db
-        .onboarded_targets()
-        .find(doc! {})
-        .skip(skip)
-        .limit(params.limit)
-        .await
-    {
-        Ok(cursor) => collect_cursor_async(cursor).await,
-        Err(e) => {
-            tracing::warn!("Failed to fetch onboarded targets: {e}");
-            Vec::new()
-        }
-    };
-    Ok(Json(ApiResponse {
-        data: targets,
-        total: Some(total),
-        page: Some(params.page),
-    }))
-}
-
-/// POST /api/v1/targets — create an onboarded target.
-#[tracing::instrument(skip_all)]
-pub async fn create_target(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Json(req): Json<CreateTargetRequest>,
-) -> Result<Json<ApiResponse<OnboardedTarget>>, StatusCode> {
-    let mut target = OnboardedTarget::new(req.name, req.target_type);
-    target.description = req.description;
-    target.artifacts = req.artifacts.iter().map(ArtifactInput::build).collect();
-
-    let db = tenant_db(&agent, &tenant).await?;
-    let res = db
-        .onboarded_targets()
-        .insert_one(&target)
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-    target.id = res.inserted_id.as_object_id();
-    Ok(Json(ApiResponse {
-        data: target,
-        total: None,
-        page: None,
-    }))
-}
-
-/// GET /api/v1/targets/{id} — fetch one target.
-#[tracing::instrument(skip_all, fields(target_id = %id))]
-pub async fn get_target(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Path(id): Path<String>,
-) -> Result<Json<ApiResponse<OnboardedTarget>>, StatusCode> {
-    let oid = parse_oid(&id)?;
-    let db = tenant_db(&agent, &tenant).await?;
-    let target = db
-        .onboarded_targets()
-        .find_one(doc! { "_id": oid })
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?
-        .ok_or(StatusCode::NOT_FOUND)?;
-    Ok(Json(ApiResponse {
-        data: target,
-        total: None,
-        page: None,
-    }))
-}
-
-/// PATCH /api/v1/targets/{id} — update mutable fields.
-#[tracing::instrument(skip_all, fields(target_id = %id))]
-pub async fn update_target(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Path(id): Path<String>,
-    Json(req): Json<UpdateTargetRequest>,
-) -> Result<Json<ApiResponse<OnboardedTarget>>, StatusCode> {
-    let oid = parse_oid(&id)?;
-    let db = tenant_db(&agent, &tenant).await?;
-
-    let mut set = doc! { "updated_at": mongodb::bson::DateTime::now() };
-    if let Some(name) = req.name {
-        set.insert("name", name);
-    }
-    if let Some(tt) = req.target_type {
-        set.insert(
-            "target_type",
-            to_bson(&tt).map_err(|_| StatusCode::BAD_REQUEST)?,
-        );
-    }
-    if let Some(sc) = req.scan_config {
-        set.insert(
-            "scan_config",
-            to_bson(&sc).map_err(|_| StatusCode::BAD_REQUEST)?,
-        );
-    }
-    if let Some(cp) = req.compliance_profile {
-        set.insert(
-            "compliance_profile",
-            to_bson(&cp).map_err(|_| StatusCode::BAD_REQUEST)?,
-        );
-    }
-    if let Some(ss) = req.scan_schedule {
-        set.insert("scan_schedule", ss);
-    }
-    if let Some(arts) = req.artifacts {
-        let built: Vec<Artifact> = arts.iter().map(ArtifactInput::build).collect();
-        set.insert(
-            "artifacts",
-            to_bson(&built).map_err(|_| StatusCode::BAD_REQUEST)?,
-        );
-    }
-
-    db.onboarded_targets()
-        .update_one(doc! { "_id": oid }, doc! { "$set": set })
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-
-    get_target(Extension(agent), tenant, Path(id)).await
-}
-
-/// DELETE /api/v1/targets/{id} — remove the target and its findings/scans.
-#[tracing::instrument(skip_all, fields(target_id = %id))]
-pub async fn delete_target(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Path(id): Path<String>,
-) -> Result<Json<serde_json::Value>, StatusCode> {
-    let oid = parse_oid(&id)?;
-    let db = tenant_db(&agent, &tenant).await?;
-    db.onboarded_targets()
-        .delete_one(doc! { "_id": oid })
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-    // Cascade all data keyed by repo_id == target id (best-effort).
-    let db = &db;
-    let _ = db.findings().delete_many(doc! { "repo_id": &id }).await;
-    let _ = db.sbom_entries().delete_many(doc! { "repo_id": &id }).await;
-    let _ = db.scan_runs().delete_many(doc! { "repo_id": &id }).await;
-    let _ = db.cve_alerts().delete_many(doc! { "repo_id": &id }).await;
-    let _ = db
-        .tracker_issues()
-        .delete_many(doc! { "repo_id": &id })
-        .await;
-    let _ = db.graph_nodes().delete_many(doc! { "repo_id": &id }).await;
-    let _ = db.graph_edges().delete_many(doc! { "repo_id": &id }).await;
-    let _ = db.graph_builds().delete_many(doc! { "repo_id": &id }).await;
-    let _ = db
-        .impact_analyses()
-        .delete_many(doc! { "repo_id": &id })
-        .await;
-    let _ = db
-        .code_embeddings()
-        .delete_many(doc! { "repo_id": &id })
-        .await;
-    let _ = db
-        .embedding_builds()
-        .delete_many(doc! { "repo_id": &id })
-        .await;
-
-    // DAST targets linked to this target, and all their downstream data.
-    if let Ok(mut cursor) = db.dast_targets().find(doc! { "repo_id": &id }).await {
-        use futures_util::StreamExt;
-        while let Some(Ok(dt)) = cursor.next().await {
-            let dast_target_id = dt.id.map(|oid| oid.to_hex()).unwrap_or_default();
-            if !dast_target_id.is_empty() {
-                cascade_delete_dast_target(db, &dast_target_id).await;
-            }
-        }
-    }
-
-    // Pentest sessions linked directly to this target (not via a DAST target).
-    if let Ok(mut cursor) = db.pentest_sessions().find(doc! { "repo_id": &id }).await {
-        use futures_util::StreamExt;
-        while let Some(Ok(session)) = cursor.next().await {
-            let session_id = session.id.map(|oid| oid.to_hex()).unwrap_or_default();
-            if !session_id.is_empty() {
-                let _ = db
-                    .attack_chain_nodes()
-                    .delete_many(doc! { "session_id": &session_id })
-                    .await;
-                let _ = db
-                    .pentest_messages()
-                    .delete_many(doc! { "session_id": &session_id })
-                    .await;
-                let _ = db
-                    .dast_findings()
-                    .delete_many(doc! { "session_id": &session_id })
-                    .await;
-            }
-        }
-    }
-    let _ = db
-        .pentest_sessions()
-        .delete_many(doc! { "repo_id": &id })
-        .await;
-
-    Ok(Json(serde_json::json!({ "status": "deleted" })))
-}
-
-/// Delete a DAST target and everything downstream of it (pentest sessions +
-/// their attack chains / messages / findings, DAST scan runs + findings).
-async fn cascade_delete_dast_target(db: &crate::database::Database, target_id: &str) {
-    use futures_util::StreamExt;
-    if let Ok(mut cursor) = db
-        .pentest_sessions()
-        .find(doc! { "target_id": target_id })
-        .await
-    {
-        while let Some(Ok(session)) = cursor.next().await {
-            let session_id = session.id.map(|oid| oid.to_hex()).unwrap_or_default();
-            if !session_id.is_empty() {
-                let _ = db
-                    .attack_chain_nodes()
-                    .delete_many(doc! { "session_id": &session_id })
-                    .await;
-                let _ = db
-                    .pentest_messages()
-                    .delete_many(doc! { "session_id": &session_id })
-                    .await;
-                let _ = db
-                    .dast_findings()
-                    .delete_many(doc! { "session_id": &session_id })
-                    .await;
-            }
-        }
-    }
-    let _ = db
-        .pentest_sessions()
-        .delete_many(doc! { "target_id": target_id })
-        .await;
-    let _ = db
-        .dast_findings()
-        .delete_many(doc! { "target_id": target_id })
-        .await;
-    let _ = db
-        .dast_scan_runs()
-        .delete_many(doc! { "target_id": target_id })
-        .await;
-    if let Ok(oid) = mongodb::bson::oid::ObjectId::parse_str(target_id) {
-        let _ = db.dast_targets().delete_one(doc! { "_id": oid }).await;
-    }
-}
-
-/// POST /api/v1/targets/{id}/artifacts — attach an artifact (by reference).
-#[tracing::instrument(skip_all, fields(target_id = %id))]
-pub async fn add_artifact(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Path(id): Path<String>,
-    Json(input): Json<ArtifactInput>,
-) -> Result<Json<ApiResponse<OnboardedTarget>>, StatusCode> {
-    let oid = parse_oid(&id)?;
-    let db = tenant_db(&agent, &tenant).await?;
-    let artifact = to_bson(&input.build()).map_err(|_| StatusCode::BAD_REQUEST)?;
-    db.onboarded_targets()
-        .update_one(
-            doc! { "_id": oid },
-            doc! { "$push": { "artifacts": artifact }, "$set": { "updated_at": mongodb::bson::DateTime::now() } },
-        )
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-    get_target(Extension(agent), tenant, Path(id)).await
-}
-
-/// POST /api/v1/targets/{id}/artifacts/upload — attach an artifact by uploading
-/// its file (PLC project, firmware image, source archive, mobile package). The
-/// bytes are written to the artifact blob store and referenced by `stored_path`,
-/// so ingest resolves them locally (no URL fetch).
-///
-/// Multipart fields: `file` (required), `kind` (required, snake_case
-/// `ArtifactKind`), `plc_format` (optional, for PLC projects).
-#[tracing::instrument(skip_all, fields(target_id = %id))]
-pub async fn upload_artifact(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Path(id): Path<String>,
-    mut multipart: Multipart,
-) -> Result<Json<ApiResponse<OnboardedTarget>>, StatusCode> {
-    let oid = parse_oid(&id)?;
-    let db = tenant_db(&agent, &tenant).await?;
-    if db
-        .onboarded_targets()
-        .find_one(doc! { "_id": oid })
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?
-        .is_none()
-    {
-        return Err(StatusCode::NOT_FOUND);
-    }
-
-    let mut kind: Option<ArtifactKind> = None;
-    let mut plc_format: Option<PlcFormat> = None;
-    let mut filename = String::from("upload.bin");
-    let mut bytes: Option<axum::body::Bytes> = None;
-
-    while let Some(field) = multipart
-        .next_field()
-        .await
-        .map_err(|_| StatusCode::BAD_REQUEST)?
-    {
-        match field.name().unwrap_or("") {
-            "kind" => {
-                let v = field.text().await.map_err(|_| StatusCode::BAD_REQUEST)?;
-                kind = parse_enum(&v);
-            }
-            "plc_format" => {
-                let v = field.text().await.map_err(|_| StatusCode::BAD_REQUEST)?;
-                plc_format = parse_enum(&v);
-            }
-            "file" => {
-                if let Some(fname) = field.file_name() {
-                    filename = fname.to_string();
-                }
-                bytes = Some(field.bytes().await.map_err(|_| StatusCode::BAD_REQUEST)?);
-            }
-            _ => {}
-        }
-    }
-
-    let (Some(kind), Some(bytes)) = (kind, bytes) else {
-        return Err(StatusCode::BAD_REQUEST);
-    };
-
-    // Store the uploaded bytes under the artifact blob store.
-    let safe_name: String = filename
-        .chars()
-        .map(|c| {
-            if c.is_ascii_alphanumeric() || matches!(c, '.' | '-' | '_') {
-                c
-            } else {
-                '_'
-            }
-        })
-        .collect();
-    let dir = std::path::Path::new(&agent.config.artifact_store_base_path)
-        .join("uploads")
-        .join(&id);
-    std::fs::create_dir_all(&dir).map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-    let dest = dir.join(format!("{}_{safe_name}", uuid::Uuid::new_v4()));
-    std::fs::write(&dest, bytes.as_ref()).map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-
-    // Build the artifact for this kind, referencing the stored file.
-    let mut artifact = match kind {
-        ArtifactKind::PlcProject => Artifact::plc_project(
-            filename.clone(),
-            plc_format.unwrap_or(PlcFormat::PlcopenXml),
-        ),
-        ArtifactKind::FirmwareImage => Artifact::firmware_image(filename.clone()),
-        ArtifactKind::SourceArchive => Artifact::source_archive(filename.clone()),
-        ArtifactKind::MobilePackage => Artifact::mobile_package(filename.clone()),
-        // Non-file kinds (git repo, live URL, container ref, text) use the JSON
-        // add-artifact endpoint, not upload.
-        _ => return Err(StatusCode::BAD_REQUEST),
-    };
-    artifact.stored_path = Some(dest.to_string_lossy().to_string());
-    artifact.size_bytes = Some(bytes.len() as u64);
-
-    let artifact_bson = to_bson(&artifact).map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-    db.onboarded_targets()
-        .update_one(
-            doc! { "_id": oid },
-            doc! { "$push": { "artifacts": artifact_bson }, "$set": { "updated_at": mongodb::bson::DateTime::now() } },
-        )
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-
-    get_target(Extension(agent), tenant, Path(id)).await
-}
-
-/// Deserialize a snake_case enum value from a plain string.
-fn parse_enum<T: for<'de> Deserialize<'de>>(s: &str) -> Option<T> {
-    serde_json::from_value(serde_json::Value::String(s.to_string())).ok()
-}
-
-/// GET /api/v1/targets/{id}/applicable-scans — the scan-applicability matrix.
-#[tracing::instrument(skip_all, fields(target_id = %id))]
-pub async fn applicable_scans_for_target(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Path(id): Path<String>,
-) -> Result<Json<ApiResponse<ApplicableScansResponse>>, StatusCode> {
-    let oid = parse_oid(&id)?;
-    let db = tenant_db(&agent, &tenant).await?;
-    let target = db
-        .onboarded_targets()
-        .find_one(doc! { "_id": oid })
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?
-        .ok_or(StatusCode::NOT_FOUND)?;
-
-    let scans = applicable_scans(&target)
-        .into_iter()
-        .map(|o| ScanOptionDto {
-            scan: o.scan.to_string(),
-            default_on: o.default_on,
-            rationale: o.rationale,
-            required_artifact: o.required_artifact.map(|k| k.to_string()),
-            blocked_reason: o.blocked_reason,
-        })
-        .collect();
-
-    Ok(Json(ApiResponse {
-        data: ApplicableScansResponse {
-            scans,
-            pentest_supported: supports_pentest(target.target_type),
-        },
-        total: None,
-        page: None,
-    }))
-}
-
-/// POST /api/v1/targets/{id}/detect — classify the target from its artifacts.
-///
-/// This is the lightweight pass: it classifies from artifact kinds without
-/// ingesting (cloning) sources, so it returns immediately. Deep detection (after
-/// ingest, with tramiton firmware analysis) is a follow-up background step.
-#[tracing::instrument(skip_all, fields(target_id = %id))]
-pub async fn detect_target(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Path(id): Path<String>,
-) -> Result<Json<ApiResponse<OnboardedTarget>>, StatusCode> {
-    let oid = parse_oid(&id)?;
-    let db = tenant_db(&agent, &tenant).await?;
-    let mut target = db
-        .onboarded_targets()
-        .find_one(doc! { "_id": oid })
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?
-        .ok_or(StatusCode::NOT_FOUND)?;
-
-    // No ingested working paths here → kind-based classification only; the mock
-    // firmware detector is never invoked (no firmware working path present).
-    let empty = HashMap::new();
-    let detector = MockFirmwareDetector { detection: None };
-    let classification = classify_target(&target, &empty, &detector)
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-
-    let classification_bson =
-        to_bson(&classification).map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-    db.onboarded_targets()
-        .update_one(
-            doc! { "_id": oid },
-            doc! { "$set": { "classification": classification_bson, "updated_at": mongodb::bson::DateTime::now() } },
-        )
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
-    target.classification = Some(classification);
-
-    Ok(Json(ApiResponse {
-        data: target,
-        total: None,
-        page: None,
-    }))
-}
-
-/// POST /api/v1/targets/{id}/scan — trigger a scan for the target.
-///
-/// Dispatches to the unified pipeline when `UNIFIED_PIPELINE` is set (else the
-/// legacy path). Runs in the background and returns immediately.
-#[tracing::instrument(skip_all, fields(target_id = %id))]
-pub async fn trigger_target_scan(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Path(id): Path<String>,
-) -> Result<Json<serde_json::Value>, StatusCode> {
-    let oid = parse_oid(&id)?;
-    let db = tenant_db(&agent, &tenant).await?;
-    // 404 if the target doesn't exist for this tenant.
-    if db
-        .onboarded_targets()
-        .find_one(doc! { "_id": oid })
-        .await
-        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?
-        .is_none()
-    {
-        return Err(StatusCode::NOT_FOUND);
-    }
-
-    let agent_clone = (*agent).clone();
-    let tenant_id = tenant.0.tenant_id.clone();
-    tokio::spawn(async move {
-        // Always the unified target pipeline — this endpoint is about an
-        // onboarded target by construction, independent of the global
-        // `unified_pipeline` transition flag used by the legacy paths.
-        if let Err(e) = agent_clone
-            .run_target_scan(
-                &tenant_id,
-                &id,
-                compliance_core::models::ScanTrigger::Manual,
-            )
-            .await
-        {
-            tracing::error!("Manual target scan failed for {id}: {e}");
-        }
-    });
-    Ok(Json(serde_json::json!({ "status": "scan_triggered" })))
-}
@@ -1,48 +0,0 @@
-//! OSCAL assessment endpoint.
-//!
-//! Returns a standard OSCAL assessment-results document for a target's findings,
-//! driven by each finding's stamped `control_refs` (from the scan's control-triage
-//! stage): mapped findings target their controls, unmapped findings are reported
-//! as-is. See `compliance_core::models::oscal_assessment`.
-
-use axum::extract::Extension;
-use axum::http::StatusCode;
-use axum::response::{IntoResponse, Response};
-use axum::Json;
-use mongodb::bson::doc;
-use serde::Deserialize;
-
-use compliance_core::models::oscal_assessment::assess;
-use compliance_core::models::Finding;
-use compliance_core::tenant_ctx::TenantCtx;
-
-use super::dto::{collect_cursor_async, tenant_db, AgentExt};
-
-#[derive(Debug, Deserialize)]
-pub struct AssessRequest {
-    /// The target / repo id whose findings are assessed.
-    pub target_id: String,
-}
-
-/// `POST /api/v1/oscal/assess` — OSCAL assessment-results for a target's findings.
-pub async fn assess_target(
-    Extension(agent): AgentExt,
-    tenant: TenantCtx,
-    Json(req): Json<AssessRequest>,
-) -> Response {
-    let db = match tenant_db(&agent, &tenant).await {
-        Ok(db) => db,
-        Err(code) => return code.into_response(),
-    };
-
-    let findings: Vec<Finding> = match db.findings().find(doc! { "repo_id": &req.target_id }).await
-    {
-        Ok(cursor) => collect_cursor_async(cursor).await,
-        Err(e) => {
-            tracing::warn!(error = %e, "failed to load findings for OSCAL assessment");
-            return StatusCode::INTERNAL_SERVER_ERROR.into_response();
-        }
-    };
-
-    Json(assess(&findings, chrono::Utc::now())).into_response()
-}
@@ -13,11 +13,10 @@ use compliance_core::models::dast::DastFinding;
 use compliance_core::models::finding::Finding;
 use compliance_core::models::pentest::*;
 use compliance_core::models::sbom::SbomEntry;
-use compliance_core::tenant_ctx::TenantCtx;

 use crate::agent::ComplianceAgent;

-use super::super::dto::{collect_cursor_async, tenant_db};
+use super::super::dto::collect_cursor_async;

 type AgentExt = Extension<Arc<ComplianceAgent>>;

@@ -36,15 +35,11 @@ pub struct ExportBody {
 #[tracing::instrument(skip_all, fields(session_id = %id))]
 pub async fn export_session_report(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
    Json(body): Json<ExportBody>,
 ) -> Result<axum::response::Response, (StatusCode, String)> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id)
        .map_err(|_| (StatusCode::BAD_REQUEST, "Invalid session ID".to_string()))?;
-    let db = tenant_db(&agent, &tenant)
-        .await
-        .map_err(|s| (s, "failed to acquire tenant database".to_string()))?;

    if body.password.len() < 8 {
        return Err((
@@ -54,7 +49,8 @@ pub async fn export_session_report(
    }

    // Fetch session
-    let session = db
+    let session = agent
+        .db
        .pentest_sessions()
        .find_one(doc! { "_id": oid })
        .await
@@ -68,7 +64,9 @@ pub async fn export_session_report(

    // Resolve target name
    let target = if let Ok(tid) = mongodb::bson::oid::ObjectId::parse_str(&session.target_id) {
-        db.dast_targets()
+        agent
+            .db
+            .dast_targets()
            .find_one(doc! { "_id": tid })
            .await
            .ok()
@@ -86,7 +84,8 @@ pub async fn export_session_report(
        .unwrap_or_default();

    // Fetch attack chain nodes
-    let nodes: Vec<AttackChainNode> = match db
+    let nodes: Vec<AttackChainNode> = match agent
+        .db
        .attack_chain_nodes()
        .find(doc! { "session_id": &id })
        .sort(doc! { "started_at": 1 })
@@ -96,8 +95,9 @@ pub async fn export_session_report(
        Err(_) => Vec::new(),
    };

-    // Fetch DAST findings for this session, then deduplicate
-    let raw_findings: Vec<DastFinding> = match db
+    // Fetch DAST findings for this session
+    let findings: Vec<DastFinding> = match agent
+        .db
        .dast_findings()
        .find(doc! { "session_id": &id })
        .sort(doc! { "severity": -1, "created_at": -1 })
@@ -106,14 +106,6 @@ pub async fn export_session_report(
        Ok(cursor) => collect_cursor_async(cursor).await,
        Err(_) => Vec::new(),
    };
-    let raw_count = raw_findings.len();
-    let findings = crate::pipeline::dedup::dedup_dast_findings(raw_findings);
-    if findings.len() < raw_count {
-        tracing::info!(
-            "Deduped DAST findings for session {id}: {raw_count} → {}",
-            findings.len()
-        );
-    }

    // Fetch SAST findings, SBOM, and code context for the linked repository
    let repo_id = session
@@ -122,7 +114,8 @@ pub async fn export_session_report(
        .or_else(|| target.as_ref().and_then(|t| t.repo_id.clone()));

    let (sast_findings, sbom_entries, code_context) = if let Some(ref rid) = repo_id {
-        let sast: Vec<Finding> = match db
+        let sast: Vec<Finding> = match agent
+            .db
            .findings()
            .find(doc! {
                "repo_id": rid,
@@ -142,7 +135,8 @@ pub async fn export_session_report(
            Err(_) => Vec::new(),
        };

-        let sbom: Vec<SbomEntry> = match db
+        let sbom: Vec<SbomEntry> = match agent
+            .db
            .sbom_entries()
            .find(doc! {
                "repo_id": rid,
@@ -162,7 +156,8 @@ pub async fn export_session_report(
        };

        // Build code context from graph nodes
-        let code_ctx: Vec<CodeContextHint> = match db
+        let code_ctx: Vec<CodeContextHint> = match agent
+            .db
            .graph_nodes()
            .find(doc! { "repo_id": rid, "is_entry_point": true })
            .limit(50)
@@ -7,12 +7,11 @@ use mongodb::bson::doc;
 use serde::Deserialize;

 use compliance_core::models::pentest::*;
-use compliance_core::tenant_ctx::TenantCtx;

 use crate::agent::ComplianceAgent;
 use crate::pentest::PentestOrchestrator;

-use super::super::dto::{collect_cursor_async, tenant_db, ApiResponse, PaginationParams};
+use super::super::dto::{collect_cursor_async, ApiResponse, PaginationParams};

 type AgentExt = Extension<Arc<ComplianceAgent>>;

@@ -44,7 +43,6 @@ pub struct LookupRepoQuery {
 #[tracing::instrument(skip_all)]
 pub async fn create_session(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Json(req): Json<CreateSessionRequest>,
 ) -> Result<Json<ApiResponse<PentestSession>>, (StatusCode, String)> {
    // Try to acquire a concurrency permit
@@ -59,10 +57,6 @@ pub async fn create_session(
            )
        })?;

-    let db = tenant_db(&agent, &tenant)
-        .await
-        .map_err(|s| (s, "failed to acquire tenant database".to_string()))?;
-
    if let Some(ref config) = req.config {
        // ── Wizard path ──────────────────────────────────────────────
        if !config.disclaimer_accepted {
@@ -73,7 +67,8 @@ pub async fn create_session(
        }

        // Look up or auto-create DastTarget by app_url
-        let target = match db
+        let target = match agent
+            .db
            .dast_targets()
            .find_one(doc! { "base_url": &config.app_url })
            .await
@@ -92,7 +87,7 @@ pub async fn create_session(
                }
                t.allow_destructive = config.allow_destructive;
                t.excluded_paths = config.scope_exclusions.clone();
-                let res = db.dast_targets().insert_one(&t).await.map_err(|e| {
+                let res = agent.db.dast_targets().insert_one(&t).await.map_err(|e| {
                    (
                        StatusCode::INTERNAL_SERVER_ERROR,
                        format!("Failed to create target: {e}"),
@@ -113,18 +108,20 @@ pub async fn create_session(
        session.config = Some(config.clone());
        session.repo_id = target.repo_id.clone();

-        // Resolve repo_id (target id) from git_repo_url if provided
+        // Resolve repo_id from git_repo_url if provided
        if let Some(ref git_url) = config.git_repo_url {
-            if let Ok(Some(target)) = db
-                .onboarded_targets()
-                .find_one(doc! { "artifacts.source_ref": git_url })
+            if let Ok(Some(repo)) = agent
+                .db
+                .repositories()
+                .find_one(doc! { "git_url": git_url })
                .await
            {
-                session.repo_id = target.id.map(|oid| oid.to_hex());
+                session.repo_id = repo.id.map(|oid| oid.to_hex());
            }
        }

-        let insert_result = db
+        let insert_result = agent
+            .db
            .pentest_sessions()
            .insert_one(&session)
            .await
@@ -215,7 +212,8 @@ pub async fn create_session(
        // Persist encrypted credentials to DB
        if session_for_task.config.is_some() {
            if let Some(sid) = session.id {
-                let _ = db
+                let _ = agent
+                    .db
                    .pentest_sessions()
                    .update_one(
                        doc! { "_id": sid },
@@ -247,13 +245,12 @@ pub async fn create_session(
            });

        let llm = agent.llm.clone();
-        let db_for_orchestrator = db.clone();
+        let db = agent.db.clone();
        let session_clone = session.clone();
        let target_clone = target.clone();
        let agent_ref = agent.clone();
        tokio::spawn(async move {
-            let orchestrator =
-                PentestOrchestrator::new(llm, db_for_orchestrator, event_tx, Some(pause_rx));
+            let orchestrator = PentestOrchestrator::new(llm, db, event_tx, Some(pause_rx));
            orchestrator
                .run_session_guarded(&session_clone, &target_clone, &initial_message)
                .await;
@@ -295,7 +292,8 @@ pub async fn create_session(
            )
        })?;

-        let target = db
+        let target = agent
+            .db
            .dast_targets()
            .find_one(doc! { "_id": oid })
            .await
@@ -312,7 +310,8 @@ pub async fn create_session(
        let mut session = PentestSession::new(target_id, strategy);
        session.repo_id = target.repo_id.clone();

-        let insert_result = db
+        let insert_result = agent
+            .db
            .pentest_sessions()
            .insert_one(&session)
            .await
@@ -339,13 +338,12 @@ pub async fn create_session(
        });

        let llm = agent.llm.clone();
-        let db_for_orchestrator = db.clone();
+        let db = agent.db.clone();
        let session_clone = session.clone();
        let target_clone = target.clone();
        let agent_ref = agent.clone();
        tokio::spawn(async move {
-            let orchestrator =
-                PentestOrchestrator::new(llm, db_for_orchestrator, event_tx, Some(pause_rx));
+            let orchestrator = PentestOrchestrator::new(llm, db, event_tx, Some(pause_rx));
            orchestrator
                .run_session_guarded(&session_clone, &target_clone, &initial_message)
                .await;
@@ -375,25 +373,21 @@ fn parse_strategy(s: &str) -> PentestStrategy {
 #[tracing::instrument(skip_all)]
 pub async fn lookup_repo(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(params): Query<LookupRepoQuery>,
 ) -> Result<Json<ApiResponse<serde_json::Value>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let repo = db
-        .onboarded_targets()
-        .find_one(doc! { "artifacts.source_ref": &params.url })
+    let repo = agent
+        .db
+        .repositories()
+        .find_one(doc! { "git_url": &params.url })
        .await
        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;

    let data = match repo {
-        Some(r) => {
-            let git = r.code_artifact().and_then(|c| c.git.as_ref());
-            serde_json::json!({
-                "name": r.name,
-                "default_branch": git.map(|g| g.default_branch.clone()),
-                "last_scanned_commit": git.and_then(|g| g.last_scanned_commit.clone()),
-            })
-        }
+        Some(r) => serde_json::json!({
+            "name": r.name,
+            "default_branch": r.default_branch,
+            "last_scanned_commit": r.last_scanned_commit,
+        }),
        None => serde_json::Value::Null,
    };

@@ -408,11 +402,9 @@ pub async fn lookup_repo(
 #[tracing::instrument(skip_all)]
 pub async fn list_sessions(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(params): Query<PaginationParams>,
 ) -> Result<Json<ApiResponse<Vec<PentestSession>>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let skip = (params.page.saturating_sub(1)) * params.limit as u64;
    let total = db
        .pentest_sessions()
@@ -446,13 +438,12 @@ pub async fn list_sessions(
 #[tracing::instrument(skip_all, fields(session_id = %id))]
 pub async fn get_session(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
 ) -> Result<Json<ApiResponse<PentestSession>>, StatusCode> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;

-    let mut session = db
+    let mut session = agent
+        .db
        .pentest_sessions()
        .find_one(doc! { "_id": oid })
        .await
@@ -480,18 +471,15 @@ pub async fn get_session(
 #[tracing::instrument(skip_all, fields(session_id = %id))]
 pub async fn send_message(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
    Json(req): Json<SendMessageRequest>,
 ) -> Result<Json<ApiResponse<PentestMessage>>, (StatusCode, String)> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id)
        .map_err(|_| (StatusCode::BAD_REQUEST, "Invalid session ID".to_string()))?;
-    let db = tenant_db(&agent, &tenant)
-        .await
-        .map_err(|s| (s, "failed to acquire tenant database".to_string()))?;

    // Verify session exists and is running
-    let session = db
+    let session = agent
+        .db
        .pentest_sessions()
        .find_one(doc! { "_id": oid })
        .await
@@ -518,7 +506,8 @@ pub async fn send_message(
        )
    })?;

-    let target = db
+    let target = agent
+        .db
        .dast_targets()
        .find_one(doc! { "_id": target_oid })
        .await
@@ -538,13 +527,13 @@ pub async fn send_message(
    // Store user message
    let session_id = id.clone();
    let user_msg = PentestMessage::user(session_id.clone(), req.message.clone());
-    let _ = db.pentest_messages().insert_one(&user_msg).await;
+    let _ = agent.db.pentest_messages().insert_one(&user_msg).await;

    let response_msg = user_msg.clone();

    // Spawn orchestrator to continue the session
    let llm = agent.llm.clone();
-    let db_for_orchestrator = db.clone();
+    let db = agent.db.clone();
    let message = req.message.clone();

    // Use existing broadcast sender if available, otherwise create a new one
@@ -559,7 +548,7 @@ pub async fn send_message(
        .unwrap_or_else(|| agent.register_session_stream(&session_id));

    tokio::spawn(async move {
-        let orchestrator = PentestOrchestrator::new(llm, db_for_orchestrator, event_tx, None);
+        let orchestrator = PentestOrchestrator::new(llm, db, event_tx, None);
        orchestrator
            .run_session_guarded(&session, &target, &message)
            .await;
@@ -576,16 +565,13 @@ pub async fn send_message(
 #[tracing::instrument(skip_all, fields(session_id = %id))]
 pub async fn stop_session(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
 ) -> Result<Json<ApiResponse<PentestSession>>, (StatusCode, String)> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id)
        .map_err(|_| (StatusCode::BAD_REQUEST, "Invalid session ID".to_string()))?;
-    let db = tenant_db(&agent, &tenant)
-        .await
-        .map_err(|s| (s, "failed to acquire tenant database".to_string()))?;

-    let session = db
+    let session = agent
+        .db
        .pentest_sessions()
        .find_one(doc! { "_id": oid })
        .await
@@ -604,7 +590,9 @@ pub async fn stop_session(
        ));
    }

-    db.pentest_sessions()
+    agent
+        .db
+        .pentest_sessions()
        .update_one(
            doc! { "_id": oid },
            doc! { "$set": {
@@ -624,7 +612,8 @@ pub async fn stop_session(
    // Clean up session resources
    agent.cleanup_session(&id);

-    let updated = db
+    let updated = agent
+        .db
        .pentest_sessions()
        .find_one(doc! { "_id": oid })
        .await
@@ -652,16 +641,13 @@ pub async fn stop_session(
 #[tracing::instrument(skip_all, fields(session_id = %id))]
 pub async fn pause_session(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
 ) -> Result<Json<ApiResponse<serde_json::Value>>, (StatusCode, String)> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id)
        .map_err(|_| (StatusCode::BAD_REQUEST, "Invalid session ID".to_string()))?;
-    let db = tenant_db(&agent, &tenant)
-        .await
-        .map_err(|s| (s, "failed to acquire tenant database".to_string()))?;

-    let session = db
+    let session = agent
+        .db
        .pentest_sessions()
        .find_one(doc! { "_id": oid })
        .await
@@ -698,16 +684,13 @@ pub async fn pause_session(
 #[tracing::instrument(skip_all, fields(session_id = %id))]
 pub async fn resume_session(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
 ) -> Result<Json<ApiResponse<serde_json::Value>>, (StatusCode, String)> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id)
        .map_err(|_| (StatusCode::BAD_REQUEST, "Invalid session ID".to_string()))?;
-    let db = tenant_db(&agent, &tenant)
-        .await
-        .map_err(|s| (s, "failed to acquire tenant database".to_string()))?;

-    let session = db
+    let session = agent
+        .db
        .pentest_sessions()
        .find_one(doc! { "_id": oid })
        .await
@@ -744,13 +727,12 @@ pub async fn resume_session(
 #[tracing::instrument(skip_all, fields(session_id = %id))]
 pub async fn get_attack_chain(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
 ) -> Result<Json<ApiResponse<Vec<AttackChainNode>>>, StatusCode> {
    let _oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;

-    let nodes = match db
+    let nodes = match agent
+        .db
        .attack_chain_nodes()
        .find(doc! { "session_id": &id })
        .sort(doc! { "started_at": 1 })
@@ -775,21 +757,21 @@ pub async fn get_attack_chain(
 #[tracing::instrument(skip_all, fields(session_id = %id))]
 pub async fn get_messages(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
    Query(params): Query<PaginationParams>,
 ) -> Result<Json<ApiResponse<Vec<PentestMessage>>>, StatusCode> {
    let _oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;

    let skip = (params.page.saturating_sub(1)) * params.limit as u64;
-    let total = db
+    let total = agent
+        .db
        .pentest_messages()
        .count_documents(doc! { "session_id": &id })
        .await
        .unwrap_or(0);

-    let messages = match db
+    let messages = match agent
+        .db
        .pentest_messages()
        .find(doc! { "session_id": &id })
        .sort(doc! { "created_at": 1 })
@@ -815,21 +797,21 @@ pub async fn get_messages(
 #[tracing::instrument(skip_all, fields(session_id = %id))]
 pub async fn get_session_findings(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
    Query(params): Query<PaginationParams>,
 ) -> Result<Json<ApiResponse<Vec<compliance_core::models::dast::DastFinding>>>, StatusCode> {
    let _oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;

    let skip = (params.page.saturating_sub(1)) * params.limit as u64;
-    let total = db
+    let total = agent
+        .db
        .dast_findings()
        .count_documents(doc! { "session_id": &id })
        .await
        .unwrap_or(0);

-    let findings = match db
+    let findings = match agent
+        .db
        .dast_findings()
        .find(doc! { "session_id": &id })
        .sort(doc! { "created_at": -1 })
@@ -6,11 +6,10 @@ use axum::Json;
 use mongodb::bson::doc;

 use compliance_core::models::pentest::*;
-use compliance_core::tenant_ctx::TenantCtx;

 use crate::agent::ComplianceAgent;

-use super::super::dto::{collect_cursor_async, tenant_db, ApiResponse};
+use super::super::dto::{collect_cursor_async, ApiResponse};

 type AgentExt = Extension<Arc<ComplianceAgent>>;

@@ -18,10 +17,8 @@ type AgentExt = Extension<Arc<ComplianceAgent>>;
 #[tracing::instrument(skip_all)]
 pub async fn pentest_stats(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
 ) -> Result<Json<ApiResponse<PentestStats>>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;

    let running_sessions = db
        .pentest_sessions()
@@ -11,11 +11,10 @@ use tokio_stream::wrappers::BroadcastStream;
 use tokio_stream::StreamExt;

 use compliance_core::models::pentest::*;
-use compliance_core::tenant_ctx::TenantCtx;

 use crate::agent::ComplianceAgent;

-use super::super::dto::{collect_cursor_async, tenant_db};
+use super::super::dto::collect_cursor_async;

 type AgentExt = Extension<Arc<ComplianceAgent>>;

@@ -26,14 +25,13 @@ type AgentExt = Extension<Arc<ComplianceAgent>>;
 #[tracing::instrument(skip_all, fields(session_id = %id))]
 pub async fn session_stream(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Path(id): Path<String>,
 ) -> Result<Sse<impl futures_util::Stream<Item = Result<Event, Infallible>>>, StatusCode> {
    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let db = tenant_db(&agent, &tenant).await?;

    // Verify session exists
-    let _session = db
+    let _session = agent
+        .db
        .pentest_sessions()
        .find_one(doc! { "_id": oid })
        .await
@@ -45,7 +43,8 @@ pub async fn session_stream(
    let mut initial_events: Vec<Result<Event, Infallible>> = Vec::new();

    // Fetch recent messages for this session
-    let messages: Vec<PentestMessage> = match db
+    let messages: Vec<PentestMessage> = match agent
+        .db
        .pentest_messages()
        .find(doc! { "session_id": &id })
        .sort(doc! { "created_at": 1 })
@@ -57,7 +56,8 @@ pub async fn session_stream(
    };

    // Fetch recent attack chain nodes
-    let nodes: Vec<AttackChainNode> = match db
+    let nodes: Vec<AttackChainNode> = match agent
+        .db
        .attack_chain_nodes()
        .find(doc! { "session_id": &id })
        .sort(doc! { "started_at": 1 })
@@ -94,7 +94,8 @@ pub async fn session_stream(
    }

    // Add current session status event
-    let session = db
+    let session = agent
+        .db
        .pentest_sessions()
        .find_one(doc! { "_id": oid })
        .await
@@ -0,0 +1,241 @@
+use axum::extract::{Extension, Path, Query};
+use axum::http::StatusCode;
+use axum::Json;
+use mongodb::bson::doc;
+
+use super::dto::*;
+use compliance_core::models::*;
+
+#[tracing::instrument(skip_all)]
+pub async fn list_repositories(
+    Extension(agent): AgentExt,
+    Query(params): Query<PaginationParams>,
+) -> ApiResult<Vec<TrackedRepository>> {
+    let db = &agent.db;
+    let skip = (params.page.saturating_sub(1)) * params.limit as u64;
+    let total = db
+        .repositories()
+        .count_documents(doc! {})
+        .await
+        .unwrap_or(0);
+
+    let repos = match db
+        .repositories()
+        .find(doc! {})
+        .skip(skip)
+        .limit(params.limit)
+        .await
+    {
+        Ok(cursor) => collect_cursor_async(cursor).await,
+        Err(e) => {
+            tracing::warn!("Failed to fetch repositories: {e}");
+            Vec::new()
+        }
+    };
+
+    Ok(Json(ApiResponse {
+        data: repos,
+        total: Some(total),
+        page: Some(params.page),
+    }))
+}
+
+#[tracing::instrument(skip_all)]
+pub async fn add_repository(
+    Extension(agent): AgentExt,
+    Json(req): Json<AddRepositoryRequest>,
+) -> Result<Json<ApiResponse<TrackedRepository>>, (StatusCode, String)> {
+    // Validate repository access before saving
+    let creds = crate::pipeline::git::RepoCredentials {
+        ssh_key_path: Some(agent.config.ssh_key_path.clone()),
+        auth_token: req.auth_token.clone(),
+        auth_username: req.auth_username.clone(),
+    };
+
+    if let Err(e) = crate::pipeline::git::GitOps::test_access(&req.git_url, &creds) {
+        return Err((
+            StatusCode::BAD_REQUEST,
+            format!("Cannot access repository: {e}"),
+        ));
+    }
+
+    let mut repo = TrackedRepository::new(req.name, req.git_url);
+    repo.default_branch = req.default_branch;
+    repo.auth_token = req.auth_token;
+    repo.auth_username = req.auth_username;
+    repo.tracker_type = req.tracker_type;
+    repo.tracker_owner = req.tracker_owner;
+    repo.tracker_repo = req.tracker_repo;
+    repo.tracker_token = req.tracker_token;
+    repo.scan_schedule = req.scan_schedule;
+
+    agent
+        .db
+        .repositories()
+        .insert_one(&repo)
+        .await
+        .map_err(|_| {
+            (
+                StatusCode::CONFLICT,
+                "Repository already exists".to_string(),
+            )
+        })?;
+
+    Ok(Json(ApiResponse {
+        data: repo,
+        total: None,
+        page: None,
+    }))
+}
+
+#[tracing::instrument(skip_all, fields(repo_id = %id))]
+pub async fn update_repository(
+    Extension(agent): AgentExt,
+    Path(id): Path<String>,
+    Json(req): Json<UpdateRepositoryRequest>,
+) -> Result<Json<serde_json::Value>, StatusCode> {
+    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
+
+    let mut set_doc = doc! { "updated_at": mongodb::bson::DateTime::now() };
+
+    if let Some(name) = &req.name {
+        set_doc.insert("name", name);
+    }
+    if let Some(branch) = &req.default_branch {
+        set_doc.insert("default_branch", branch);
+    }
+    if let Some(token) = &req.auth_token {
+        set_doc.insert("auth_token", token);
+    }
+    if let Some(username) = &req.auth_username {
+        set_doc.insert("auth_username", username);
+    }
+    if let Some(tracker_type) = &req.tracker_type {
+        set_doc.insert("tracker_type", tracker_type.to_string());
+    }
+    if let Some(owner) = &req.tracker_owner {
+        set_doc.insert("tracker_owner", owner);
+    }
+    if let Some(repo) = &req.tracker_repo {
+        set_doc.insert("tracker_repo", repo);
+    }
+    if let Some(token) = &req.tracker_token {
+        set_doc.insert("tracker_token", token);
+    }
+    if let Some(schedule) = &req.scan_schedule {
+        set_doc.insert("scan_schedule", schedule);
+    }
+
+    let result = agent
+        .db
+        .repositories()
+        .update_one(doc! { "_id": oid }, doc! { "$set": set_doc })
+        .await
+        .map_err(|e| {
+            tracing::warn!("Failed to update repository: {e}");
+            StatusCode::INTERNAL_SERVER_ERROR
+        })?;
+
+    if result.matched_count == 0 {
+        return Err(StatusCode::NOT_FOUND);
+    }
+
+    Ok(Json(serde_json::json!({ "status": "updated" })))
+}
+
+#[tracing::instrument(skip_all)]
+pub async fn get_ssh_public_key(
+    Extension(agent): AgentExt,
+) -> Result<Json<serde_json::Value>, StatusCode> {
+    let public_path = format!("{}.pub", agent.config.ssh_key_path);
+    let public_key = std::fs::read_to_string(&public_path).map_err(|_| StatusCode::NOT_FOUND)?;
+    Ok(Json(serde_json::json!({ "public_key": public_key.trim() })))
+}
+
+#[tracing::instrument(skip_all, fields(repo_id = %id))]
+pub async fn trigger_scan(
+    Extension(agent): AgentExt,
+    Path(id): Path<String>,
+) -> Result<Json<serde_json::Value>, StatusCode> {
+    let agent_clone = (*agent).clone();
+    tokio::spawn(async move {
+        if let Err(e) = agent_clone.run_scan(&id, ScanTrigger::Manual).await {
+            tracing::error!("Manual scan failed for {id}: {e}");
+        }
+    });
+
+    Ok(Json(serde_json::json!({ "status": "scan_triggered" })))
+}
+
+/// Return the webhook secret for a repository (used by dashboard to display it)
+pub async fn get_webhook_config(
+    Extension(agent): AgentExt,
+    Path(id): Path<String>,
+) -> Result<Json<serde_json::Value>, StatusCode> {
+    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
+    let repo = agent
+        .db
+        .repositories()
+        .find_one(doc! { "_id": oid })
+        .await
+        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?
+        .ok_or(StatusCode::NOT_FOUND)?;
+
+    let tracker_type = repo
+        .tracker_type
+        .as_ref()
+        .map(|t| t.to_string())
+        .unwrap_or_else(|| "gitea".to_string());
+
+    Ok(Json(serde_json::json!({
+        "webhook_secret": repo.webhook_secret,
+        "tracker_type": tracker_type,
+    })))
+}
+
+#[tracing::instrument(skip_all, fields(repo_id = %id))]
+pub async fn delete_repository(
+    Extension(agent): AgentExt,
+    Path(id): Path<String>,
+) -> Result<Json<serde_json::Value>, StatusCode> {
+    let oid = mongodb::bson::oid::ObjectId::parse_str(&id).map_err(|_| StatusCode::BAD_REQUEST)?;
+    let db = &agent.db;
+
+    // Delete the repository
+    let result = db
+        .repositories()
+        .delete_one(doc! { "_id": oid })
+        .await
+        .map_err(|_| StatusCode::INTERNAL_SERVER_ERROR)?;
+
+    if result.deleted_count == 0 {
+        return Err(StatusCode::NOT_FOUND);
+    }
+
+    // Cascade delete all related data
+    let _ = db.findings().delete_many(doc! { "repo_id": &id }).await;
+    let _ = db.sbom_entries().delete_many(doc! { "repo_id": &id }).await;
+    let _ = db.scan_runs().delete_many(doc! { "repo_id": &id }).await;
+    let _ = db.cve_alerts().delete_many(doc! { "repo_id": &id }).await;
+    let _ = db
+        .tracker_issues()
+        .delete_many(doc! { "repo_id": &id })
+        .await;
+    let _ = db.graph_nodes().delete_many(doc! { "repo_id": &id }).await;
+    let _ = db.graph_edges().delete_many(doc! { "repo_id": &id }).await;
+    let _ = db.graph_builds().delete_many(doc! { "repo_id": &id }).await;
+    let _ = db
+        .impact_analyses()
+        .delete_many(doc! { "repo_id": &id })
+        .await;
+    let _ = db
+        .code_embeddings()
+        .delete_many(doc! { "repo_id": &id })
+        .await;
+    let _ = db
+        .embedding_builds()
+        .delete_many(doc! { "repo_id": &id })
+        .await;
+
+    Ok(Json(serde_json::json!({ "status": "deleted" })))
+}
@@ -6,7 +6,6 @@ use mongodb::bson::doc;

 use super::dto::*;
 use compliance_core::models::SbomEntry;
-use compliance_core::tenant_ctx::TenantCtx;

 const COPYLEFT_LICENSES: &[&str] = &[
    "GPL-2.0",
@@ -30,10 +29,8 @@ const COPYLEFT_LICENSES: &[&str] = &[
 #[tracing::instrument(skip_all)]
 pub async fn sbom_filters(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
 ) -> Result<Json<serde_json::Value>, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;

    let managers: Vec<String> = db
        .sbom_entries()
@@ -64,11 +61,9 @@ pub async fn sbom_filters(
 #[tracing::instrument(skip_all, fields(repo_id = ?filter.repo_id, package_manager = ?filter.package_manager))]
 pub async fn list_sbom(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(filter): Query<SbomFilter>,
 ) -> ApiResult<Vec<SbomEntry>> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let mut query = doc! {};

    if let Some(repo_id) = &filter.repo_id {
@@ -125,11 +120,9 @@ pub async fn list_sbom(
 #[tracing::instrument(skip_all)]
 pub async fn export_sbom(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(params): Query<SbomExportParams>,
 ) -> Result<impl IntoResponse, StatusCode> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let entries: Vec<SbomEntry> = match db
        .sbom_entries()
        .find(doc! { "repo_id": &params.repo_id })
@@ -243,11 +236,9 @@ pub async fn export_sbom(
 #[tracing::instrument(skip_all)]
 pub async fn license_summary(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(params): Query<SbomFilter>,
 ) -> ApiResult<Vec<LicenseSummary>> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let mut query = doc! {};
    if let Some(repo_id) = &params.repo_id {
        query.insert("repo_id", repo_id);
@@ -282,7 +273,7 @@ pub async fn license_summary(
            }
        })
        .collect();
-    summaries.sort_by_key(|s| std::cmp::Reverse(s.count));
+    summaries.sort_by(|a, b| b.count.cmp(&a.count));

    Ok(Json(ApiResponse {
        data: summaries,
@@ -294,11 +285,9 @@ pub async fn license_summary(
 #[tracing::instrument(skip_all)]
 pub async fn sbom_diff(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(params): Query<SbomDiffParams>,
 ) -> ApiResult<SbomDiffResult> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;

    let entries_a: Vec<SbomEntry> = match db
        .sbom_entries()
@@ -4,16 +4,13 @@ use mongodb::bson::doc;

 use super::dto::*;
 use compliance_core::models::ScanRun;
-use compliance_core::tenant_ctx::TenantCtx;

 #[tracing::instrument(skip_all)]
 pub async fn list_scan_runs(
    Extension(agent): AgentExt,
-    tenant: TenantCtx,
    Query(params): Query<PaginationParams>,
 ) -> ApiResult<Vec<ScanRun>> {
-    let db = tenant_db(&agent, &tenant).await?;
-    let db = &db;
+    let db = &agent.db;
    let skip = (params.page.saturating_sub(1)) * params.limit as u64;
    let total = db.scan_runs().count_documents(doc! {}).await.unwrap_or(0);

@@ -1,289 +0,0 @@
-//! Werkbank runner endpoints (`/api/v1/werkbank/jobs/*`).
-//!
-//! The pull API a Werkbank runner talks to: lease a job, heartbeat while it runs,
-//! and post the result back. Machine auth is a **static bearer token**
-//! (`WERKBANK_RUNNER_TOKEN`) — not a Keycloak JWT, because a runner acts across
-//! tenants (each request names its `tenant`). Routes are only mounted when the
-//! token is configured; with none set they don't exist (404).
-//!
-//! On completion the runner's findings are persisted against the job's target,
-//! so a job run by a remote runner lands the same findings an in-process run
-//! would (WB-05, the control-plane cut-over).
-
-use axum::extract::{Extension, Path, Request};
-use axum::http::{header, StatusCode};
-use axum::middleware::Next;
-use axum::response::{IntoResponse, Response};
-use axum::Json;
-use mongodb::bson::{doc, oid::ObjectId};
-use secrecy::ExposeSecret;
-use serde::{Deserialize, Serialize};
-use std::time::Duration;
-
-use compliance_core::models::werkbank::{
-    CompleteRequest, CompleteResponse, HeartbeatRequest, InputRef, Job, JobResult, LeaseRequest,
-};
-use compliance_core::models::ArtifactKind;
-
-use super::dto::AgentExt;
-use crate::database::Database;
-use crate::werkbank::JobQueue;
-
-/// Gate the runner endpoints behind the static runner bearer token.
-pub async fn require_runner_token(
-    Extension(agent): AgentExt,
-    request: Request,
-    next: Next,
-) -> Response {
-    let Some(expected) = agent.config.werkbank_runner_token.as_ref() else {
-        return (StatusCode::NOT_FOUND, "werkbank runner API disabled").into_response();
-    };
-    let presented = request
-        .headers()
-        .get(header::AUTHORIZATION)
-        .and_then(|v| v.to_str().ok())
-        .and_then(|s| s.strip_prefix("Bearer "))
-        .map(str::trim)
-        .filter(|s| !s.is_empty());
-    let Some(presented) = presented else {
-        return (StatusCode::UNAUTHORIZED, "Missing bearer token").into_response();
-    };
-    if !constant_time_eq(presented, expected.expose_secret()) {
-        return (StatusCode::UNAUTHORIZED, "Invalid runner token").into_response();
-    }
-    next.run(request).await
-}
-
-/// `POST /api/v1/werkbank/jobs/lease` — lease the oldest runnable job, or `204`.
-#[tracing::instrument(skip_all, fields(tenant = %req.tenant, runner = %req.runner_id))]
-pub async fn lease(
-    Extension(agent): AgentExt,
-    Json(req): Json<LeaseRequest>,
-) -> Result<Response, StatusCode> {
-    let queue = JobQueue::new(&tenant_db(&agent, &req.tenant).await?);
-    let leased = queue
-        .lease(
-            &req.runner_id,
-            req.executor,
-            &req.labels,
-            Duration::from_secs(req.lease_ttl_secs),
-            chrono::Utc::now(),
-        )
-        .await
-        .map_err(internal)?;
-    Ok(match leased {
-        Some(job) => Json(job).into_response(),
-        None => StatusCode::NO_CONTENT.into_response(),
-    })
-}
-
-/// `POST /api/v1/werkbank/jobs/heartbeat` — extend the lease; `409` if it's lost.
-#[tracing::instrument(skip_all, fields(tenant = %req.tenant, job = %req.job_id))]
-pub async fn heartbeat(
-    Extension(agent): AgentExt,
-    Json(req): Json<HeartbeatRequest>,
-) -> Result<Response, StatusCode> {
-    let queue = JobQueue::new(&tenant_db(&agent, &req.tenant).await?);
-    let ack = queue
-        .heartbeat(
-            &req.job_id,
-            &req.lease_token,
-            Duration::from_secs(req.lease_ttl_secs),
-            chrono::Utc::now(),
-        )
-        .await
-        .map_err(internal)?;
-    Ok(match ack {
-        Some(ack) => Json(ack).into_response(),
-        // Lease lost — the runner should abandon the job.
-        None => StatusCode::CONFLICT.into_response(),
-    })
-}
-
-/// `POST /api/v1/werkbank/jobs/complete` — record the result and persist findings.
-#[tracing::instrument(skip_all, fields(tenant = %req.tenant, job = %req.job_id))]
-pub async fn complete(
-    Extension(agent): AgentExt,
-    Json(req): Json<CompleteRequest>,
-) -> Result<Json<CompleteResponse>, StatusCode> {
-    let db = tenant_db(&agent, &req.tenant).await?;
-    let queue = JobQueue::new(&db);
-    let now = chrono::Utc::now();
-    let recorded = queue
-        .complete(&req.job_id, &req.lease_token, &req.result, now)
-        .await
-        .map_err(internal)?;
-
-    // Only persist findings for the run that actually recorded the result, so a
-    // duplicate/late completion can't double-insert.
-    if recorded {
-        if let Some(record) = queue.get(&req.job_id).await.map_err(internal)? {
-            persist_findings(&db, &record.job.target_id, &req.result).await;
-        }
-    }
-    Ok(Json(CompleteResponse { recorded }))
-}
-
-/// `GET /api/v1/werkbank/artifacts/{hash}` — serve a content-addressed blob (the
-/// program a runner needs to load). The hash is validated against traversal by
-/// [`crate::ingest::blob::read_blob`]; a runner fetches this for a job's `blob`
-/// input.
-#[tracing::instrument(skip_all, fields(hash = %hash))]
-pub async fn serve_artifact(
-    Extension(agent): AgentExt,
-    Path(hash): Path<String>,
-) -> Result<Response, StatusCode> {
-    let base = std::path::Path::new(&agent.config.artifact_store_base_path);
-    match crate::ingest::blob::read_blob(base, &hash) {
-        Ok(bytes) => {
-            Ok(([(header::CONTENT_TYPE, "application/octet-stream")], bytes).into_response())
-        }
-        Err(_) => Err(StatusCode::NOT_FOUND),
-    }
-}
-
-/// Enqueue a `plc-provision` job for a target: extract its control-logic program,
-/// stash it as a content-addressed blob (which the runner fetches via
-/// [`serve_artifact`]), and queue the job. This is the control-plane "enqueue"
-/// half of the loop — a runner then leases it, provisions, and posts results.
-#[derive(Debug, Deserialize)]
-pub struct EnqueueRequest {
-    /// The tenant whose queue to enqueue into.
-    pub tenant: String,
-    /// The onboarded target to test.
-    pub target_id: String,
-}
-
-/// The enqueued job's id.
-#[derive(Debug, Serialize)]
-pub struct EnqueueResponse {
-    /// The new job id.
-    pub job_id: String,
-    /// Whether this call inserted it (false = already queued).
-    pub enqueued: bool,
-}
-
-#[tracing::instrument(skip_all, fields(tenant = %req.tenant, target = %req.target_id))]
-pub async fn enqueue(
-    Extension(agent): AgentExt,
-    Json(req): Json<EnqueueRequest>,
-) -> Result<Json<EnqueueResponse>, StatusCode> {
-    let db = tenant_db(&agent, &req.tenant).await?;
-    let oid = ObjectId::parse_str(&req.target_id).map_err(|_| StatusCode::BAD_REQUEST)?;
-    let target = db
-        .onboarded_targets()
-        .find_one(doc! { "_id": oid })
-        .await
-        .map_err(internal)?
-        .ok_or(StatusCode::NOT_FOUND)?;
-
-    // Extract the control-logic program from the target's PLC-source artifacts
-    // (same selection as the in-process PLC scan).
-    let ctx = crate::ingest::IngestContext::from_config(&agent.config, &req.target_id);
-    let ingest_set = crate::ingest::ingest_all(&target, &ctx).map_err(internal)?;
-    let program = target
-        .artifacts
-        .iter()
-        .filter(|a| {
-            matches!(
-                a.kind,
-                ArtifactKind::PlcProject | ArtifactKind::GitRepo | ArtifactKind::SourceArchive
-            )
-        })
-        .find_map(|a| {
-            let path = ingest_set
-                .get(&a.id)
-                .and_then(|ia| ia.working_path.clone())?;
-            werkbank_exec::plc::extract_program(&path)
-        })
-        .ok_or(StatusCode::UNPROCESSABLE_ENTITY)?;
-
-    // Stash the program source so the runner can fetch it by hash.
-    let base = std::path::Path::new(&agent.config.artifact_store_base_path);
-    let hash =
-        crate::ingest::blob::store_bytes(base, program.source.as_bytes()).map_err(internal)?;
-
-    let job_id = format!("job_{}", uuid::Uuid::new_v4().simple());
-    let job = Job::plc_provision(
-        &job_id,
-        &req.tenant,
-        &req.target_id,
-        InputRef::blob(hash),
-        agent.config.plc_runtime.max_lifetime_secs,
-    );
-    let enqueued = JobQueue::new(&db)
-        .enqueue(job, chrono::Utc::now())
-        .await
-        .map_err(internal)?;
-    Ok(Json(EnqueueResponse { job_id, enqueued }))
-}
-
-/// Persist a job result's findings against its target: general findings
-/// (dedup'd by fingerprint) and DAST findings. Best-effort — a persistence hiccup
-/// is logged, not surfaced to the runner (its result is already recorded).
-async fn persist_findings(db: &Database, target_id: &str, result: &JobResult) {
-    for finding in &result.findings {
-        let exists = db
-            .findings()
-            .find_one(doc! { "fingerprint": &finding.fingerprint })
-            .await
-            .ok()
-            .flatten()
-            .is_some();
-        if !exists {
-            if let Err(e) = db.findings().insert_one(finding).await {
-                tracing::warn!(target_id, error = %e, "werkbank: persist finding failed");
-            }
-        }
-    }
-    for finding in &result.dast_findings {
-        if let Err(e) = db.dast_findings().insert_one(finding).await {
-            tracing::warn!(target_id, error = %e, "werkbank: persist DAST finding failed");
-        }
-    }
-    tracing::info!(
-        target_id,
-        findings = result.findings.len(),
-        dast = result.dast_findings.len(),
-        "werkbank: persisted runner results"
-    );
-}
-
-/// Resolve the tenant-scoped database for a request.
-async fn tenant_db(
-    agent: &crate::agent::ComplianceAgent,
-    tenant: &str,
-) -> Result<Database, StatusCode> {
-    agent.db_pool.for_tenant_id(tenant).await.map_err(internal)
-}
-
-/// Map any internal error to a 500.
-fn internal<E: std::fmt::Display>(e: E) -> StatusCode {
-    tracing::error!("werkbank endpoint error: {e}");
-    StatusCode::INTERNAL_SERVER_ERROR
-}
-
-/// Length-checked, constant-time-ish token comparison.
-fn constant_time_eq(a: &str, b: &str) -> bool {
-    if a.len() != b.len() {
-        return false;
-    }
-    let mut diff = 0u8;
-    for (x, y) in a.bytes().zip(b.bytes()) {
-        diff |= x ^ y;
-    }
-    diff == 0
-}
-
-#[cfg(test)]
-mod tests {
-    use super::constant_time_eq;
-
-    #[test]
-    fn token_compare() {
-        assert!(constant_time_eq("secret", "secret"));
-        assert!(!constant_time_eq("secret", "secrex"));
-        assert!(!constant_time_eq("secret", "secretx"));
-        assert!(!constant_time_eq("", "x"));
-    }
-}
@@ -1,3 +1,4 @@
+pub mod auth_middleware;
 pub mod handlers;
 pub mod routes;
 pub mod server;
@@ -2,46 +2,29 @@ use axum::routing::{delete, get, patch, post};
 use axum::Router;

 use crate::api::handlers;
+use crate::webhooks;

 pub fn build_router() -> Router {
    Router::new()
        .route("/api/v1/health", get(handlers::health))
-        .route("/api/v1/oscal/assess", post(handlers::oscal::assess_target))
        .route("/api/v1/stats/overview", get(handlers::stats_overview))
        .route(
            "/api/v1/settings/ssh-public-key",
            get(handlers::get_ssh_public_key),
        )
-        // Unified onboarding targets (#131).
+        .route("/api/v1/repositories", get(handlers::list_repositories))
+        .route("/api/v1/repositories", post(handlers::add_repository))
        .route(
-            "/api/v1/targets",
-            get(handlers::onboarding::list_targets).post(handlers::onboarding::create_target),
+            "/api/v1/repositories/{id}/scan",
+            post(handlers::trigger_scan),
        )
        .route(
-            "/api/v1/targets/{id}",
-            get(handlers::onboarding::get_target)
-                .patch(handlers::onboarding::update_target)
-                .delete(handlers::onboarding::delete_target),
+            "/api/v1/repositories/{id}",
+            delete(handlers::delete_repository).patch(handlers::update_repository),
        )
        .route(
-            "/api/v1/targets/{id}/artifacts",
-            post(handlers::onboarding::add_artifact),
-        )
-        .route(
-            "/api/v1/targets/{id}/artifacts/upload",
-            post(handlers::onboarding::upload_artifact),
-        )
-        .route(
-            "/api/v1/targets/{id}/applicable-scans",
-            get(handlers::onboarding::applicable_scans_for_target),
-        )
-        .route(
-            "/api/v1/targets/{id}/detect",
-            post(handlers::onboarding::detect_target),
-        )
-        .route(
-            "/api/v1/targets/{id}/scan",
-            post(handlers::onboarding::trigger_target_scan),
+            "/api/v1/repositories/{id}/webhook-config",
+            get(handlers::get_webhook_config),
        )
        .route("/api/v1/findings", get(handlers::list_findings))
        .route("/api/v1/findings/{id}", get(handlers::get_finding))
@@ -64,15 +47,6 @@ pub fn build_router() -> Router {
        .route("/api/v1/sbom/diff", get(handlers::sbom_diff))
        .route("/api/v1/issues", get(handlers::list_issues))
        .route("/api/v1/scan-runs", get(handlers::list_scan_runs))
-        // MCP token management (per-tenant API tokens for the MCP server)
-        .route(
-            "/api/v1/mcp-tokens",
-            get(handlers::mcp_tokens::list_mcp_tokens).post(handlers::mcp_tokens::create_mcp_token),
-        )
-        .route(
-            "/api/v1/mcp-tokens/{id}",
-            delete(handlers::mcp_tokens::revoke_mcp_token),
-        )
        // Graph API endpoints
        .route("/api/v1/graph/{repo_id}", get(handlers::graph::get_graph))
        .route(
@@ -125,29 +99,6 @@ pub fn build_router() -> Router {
            "/api/v1/chat/{repo_id}/status",
            get(handlers::chat::embedding_status),
        )
-        // Help chat (documentation-grounded Q&A)
-        .route("/api/v1/help/chat", post(handlers::help_chat::help_chat))
-        // CVE notification endpoints
-        .route(
-            "/api/v1/notifications",
-            get(handlers::notifications::list_notifications),
-        )
-        .route(
-            "/api/v1/notifications/count",
-            get(handlers::notifications::notification_count),
-        )
-        .route(
-            "/api/v1/notifications/read-all",
-            post(handlers::notifications::mark_all_read),
-        )
-        .route(
-            "/api/v1/notifications/{id}/read",
-            patch(handlers::notifications::mark_read),
-        )
-        .route(
-            "/api/v1/notifications/{id}/dismiss",
-            patch(handlers::notifications::dismiss_notification),
-        )
        // Pentest API endpoints
        .route(
            "/api/v1/pentest/lookup-repo",
@@ -201,10 +152,17 @@ pub fn build_router() -> Router {
            "/api/v1/pentest/stats",
            get(handlers::pentest::pentest_stats),
        )
-    // Webhook routes live on the separate webhook server (port 3002,
-    // see crate::webhooks::server). The M7.2-C tenant-in-URL form is
-    // `/webhook/{tenant_id}/{platform}/{repo_id}` and the handlers
-    // expect a (tenant_id, repo_id) path tuple. Anything mounting
-    // them here on the API server would mismatch the handler
-    // signature, so the routes are not exported.
+        // Webhook endpoints (proxied through dashboard)
+        .route(
+            "/webhook/github/{repo_id}",
+            post(webhooks::github::handle_github_webhook),
+        )
+        .route(
+            "/webhook/gitlab/{repo_id}",
+            post(webhooks::gitlab::handle_gitlab_webhook),
+        )
+        .route(
+            "/webhook/gitea/{repo_id}",
+            post(webhooks::gitea::handle_gitea_webhook),
+        )
 }
@@ -1,137 +1,20 @@
 use std::sync::Arc;

-use axum::extract::{DefaultBodyLimit, Request};
-use axum::http::HeaderValue;
-use axum::middleware::Next;
-use axum::response::Response;
-use axum::routing::{delete, get, post};
-use axum::{middleware, Extension, Router};
+use axum::{middleware, Extension};
 use tokio::sync::RwLock;
 use tower_http::cors::CorsLayer;
-use tower_http::set_header::SetResponseHeaderLayer;
 use tower_http::trace::TraceLayer;

-use compliance_core::auth::{require_jwt_auth, require_tenant_status, JwksState};
-use compliance_core::{TenantContext, TenantStatus};
-
 use crate::agent::ComplianceAgent;
-use crate::api::handlers;
+use crate::api::auth_middleware::{require_jwt_auth, JwksState};
 use crate::api::routes;
 use crate::error::AgentError;

-/// Synthetic tenant id used when Keycloak isn't configured (local dev,
-/// `cargo run` against a bare Mongo). Lets the handler stack stay
-/// uniformly tenant-scoped without the operator having to spin up KC
-/// just to poke at the API. Override via `DEV_TENANT_ID`.
-const DEFAULT_DEV_TENANT_ID: &str = "dev";
-
-/// Inject a synthetic [`TenantContext`] for any request that lacks one.
-/// Only mounted when Keycloak is NOT configured; with KC, the real
-/// `require_jwt_auth` middleware owns this and we never reach here
-/// without a context.
-///
-/// Public so the integration-test harness can mount it without
-/// duplicating the synthetic-context shape.
-pub async fn inject_dev_tenant(mut request: Request, next: Next) -> Response {
-    if request.extensions().get::<TenantContext>().is_none() {
-        let tenant_id =
-            std::env::var("DEV_TENANT_ID").unwrap_or_else(|_| DEFAULT_DEV_TENANT_ID.to_string());
-        let ctx = TenantContext {
-            tenant_slug: tenant_id.clone(),
-            tenant_id,
-            org_roles: vec![],
-            products: vec![],
-            plan: "dev".to_string(),
-            status: TenantStatus::Active,
-            user_id: "dev-user".to_string(),
-            user_name: None,
-        };
-        request.extensions_mut().insert(ctx);
-    }
-    next.run(request).await
-}
-
 pub async fn start_api_server(agent: ComplianceAgent, port: u16) -> Result<(), AgentError> {
-    // Admin sub-router. Routes are only mounted when ADMIN_API_TOKEN is
-    // configured — without it, the paths don't exist at all (404 rather
-    // than 401), so an operator who hasn't opted in can't fingerprint
-    // the surface area.
-    let admin_router: Router = if agent.config.admin_api_token.is_some() {
-        tracing::info!("Admin API enabled — /api/v1/admin/* mounted behind ADMIN_API_TOKEN bearer");
-        Router::new()
-            .route(
-                "/api/v1/admin/tenants",
-                get(handlers::admin::list_tenant_dbs),
-            )
-            .route(
-                "/api/v1/admin/tenants/{tenant_id}",
-                delete(handlers::admin::drop_tenant_db),
-            )
-            .layer(middleware::from_fn(handlers::admin::require_admin_token))
-    } else {
-        Router::new()
-    };
-
-    // Werkbank runner API. Like admin, only mounted when its bearer token is
-    // configured; runners authenticate with WERKBANK_RUNNER_TOKEN (not a JWT).
-    let werkbank_router: Router = if agent.config.werkbank_runner_token.is_some() {
-        tracing::info!(
-            "Werkbank runner API enabled — /api/v1/werkbank/jobs/* behind WERKBANK_RUNNER_TOKEN"
-        );
-        Router::new()
-            .route(
-                "/api/v1/werkbank/jobs/lease",
-                post(handlers::werkbank_jobs::lease),
-            )
-            .route(
-                "/api/v1/werkbank/jobs/heartbeat",
-                post(handlers::werkbank_jobs::heartbeat),
-            )
-            .route(
-                "/api/v1/werkbank/jobs/complete",
-                post(handlers::werkbank_jobs::complete),
-            )
-            .route(
-                "/api/v1/werkbank/jobs/enqueue",
-                post(handlers::werkbank_jobs::enqueue),
-            )
-            .route(
-                "/api/v1/werkbank/artifacts/{hash}",
-                get(handlers::werkbank_jobs::serve_artifact),
-            )
-            .layer(middleware::from_fn(
-                handlers::werkbank_jobs::require_runner_token,
-            ))
-    } else {
-        Router::new()
-    };
-
    let mut app = routes::build_router()
-        .merge(admin_router)
-        .merge(werkbank_router)
-        // Allow large artifact uploads (PLC .projectarchive, firmware images,
-        // mobile packages) — axum's default request-body limit is only 2 MiB.
-        .layer(DefaultBodyLimit::max(512 * 1024 * 1024))
        .layer(Extension(Arc::new(agent.clone())))
        .layer(CorsLayer::permissive())
-        .layer(TraceLayer::new_for_http())
-        // Security headers (defense-in-depth, primary enforcement via Traefik)
-        .layer(SetResponseHeaderLayer::overriding(
-            axum::http::header::STRICT_TRANSPORT_SECURITY,
-            HeaderValue::from_static("max-age=31536000; includeSubDomains"),
-        ))
-        .layer(SetResponseHeaderLayer::overriding(
-            axum::http::header::X_FRAME_OPTIONS,
-            HeaderValue::from_static("DENY"),
-        ))
-        .layer(SetResponseHeaderLayer::overriding(
-            axum::http::header::X_CONTENT_TYPE_OPTIONS,
-            HeaderValue::from_static("nosniff"),
-        ))
-        .layer(SetResponseHeaderLayer::overriding(
-            axum::http::header::REFERRER_POLICY,
-            HeaderValue::from_static("strict-origin-when-cross-origin"),
-        ));
+        .layer(TraceLayer::new_for_http());

    if let (Some(kc_url), Some(kc_realm)) =
        (&agent.config.keycloak_url, &agent.config.keycloak_realm)
@@ -142,22 +25,11 @@ pub async fn start_api_server(agent: ComplianceAgent, port: u16) -> Result<(), A
            jwks_url,
        };
        tracing::info!("Keycloak JWT auth enabled for realm '{kc_realm}'");
-        // Layers execute outermost-first. Extension(jwks_state) must run
-        // before require_jwt_auth so the middleware can read it; the
-        // status gate runs after JWT so TenantContext is in extensions.
        app = app
-            .layer(middleware::from_fn(require_tenant_status))
-            .layer(middleware::from_fn(require_jwt_auth))
-            .layer(Extension(jwks_state));
+            .layer(Extension(jwks_state))
+            .layer(middleware::from_fn(require_jwt_auth));
    } else {
-        let tenant_id =
-            std::env::var("DEV_TENANT_ID").unwrap_or_else(|_| DEFAULT_DEV_TENANT_ID.to_string());
-        tracing::warn!(
-            tenant_id = %tenant_id,
-            "Keycloak not configured — running unauthenticated against the dev tenant. \
-             DO NOT use in any environment with real customer data."
-        );
-        app = app.layer(middleware::from_fn(inject_dev_tenant));
+        tracing::warn!("Keycloak not configured - API endpoints are unprotected");
    }

    let addr = format!("0.0.0.0:{port}");
@@ -1,217 +0,0 @@
-//! Firmware classification via tramiton.
-//!
-//! tramiton is the company's firmware build/repro engine; we do not re-implement
-//! its detection. We depend on `tramiton-core` directly (same-company IP) and run
-//! its provider analysis in-process behind a [`FirmwareDetector`] port, mapping
-//! tramiton's `BuildPlan` onto a [`TargetType`]. A deterministic
-//! [`MockFirmwareDetector`] backs the tests so CI unit tests need neither the
-//! tramiton sources nor a real firmware tree.
-
-use std::path::Path;
-
-use compliance_core::error::CoreError;
-use compliance_core::models::{DetectedFact, TargetType};
-use compliance_core::traits::ClassifierVerdict;
-
-/// A minimal firmware-detection summary, mapped from tramiton's `BuildPlan`.
-/// Kept small and tramiton-independent so the classifier and the test mock don't
-/// need to construct a full tramiton plan.
-#[derive(Debug, Clone, Default)]
-pub struct FirmwareDetection {
-    /// The detecting provider (e.g. `zephyr`, `cmake`, `source-archaeology`).
-    pub provider: String,
-    /// Detection confidence: `low` | `medium` | `high`.
-    pub confidence: String,
-    /// Build-system label (e.g. `Zephyr`, `ESP-IDF`, `CMake`).
-    pub build_system: String,
-    /// Framework, when known (`zephyr`, `esp-idf`, `bare-metal`, ...).
-    pub framework: Option<String>,
-    /// Target board / MCU / arch.
-    pub target: FirmwareTarget,
-    /// Unresolved gaps in the plan.
-    pub gaps: Vec<String>,
-}
-
-/// The detected firmware target (board / MCU / arch).
-#[derive(Debug, Clone, Default)]
-pub struct FirmwareTarget {
-    /// Board name.
-    pub board: Option<String>,
-    /// MCU part.
-    pub mcu: Option<String>,
-    /// Architecture.
-    pub arch: Option<String>,
-}
-
-/// A source of tramiton firmware detection.
-#[allow(async_fn_in_trait)]
-pub trait FirmwareDetector: Send + Sync {
-    /// Run detection over a path, returning a firmware detection if tramiton
-    /// could form a build plan.
-    async fn detect(&self, path: &Path) -> Result<Option<FirmwareDetection>, CoreError>;
-}
-
-/// Uses `tramiton-core` in-process. The analysis is blocking (filesystem walk),
-/// so it runs on a blocking thread to avoid stalling the async runtime. A path
-/// with no recognizable build system yields `Ok(None)`.
-pub struct TramitonNative;
-
-impl FirmwareDetector for TramitonNative {
-    async fn detect(&self, path: &Path) -> Result<Option<FirmwareDetection>, CoreError> {
-        let path = path.to_path_buf();
-        let plan = tokio::task::spawn_blocking(move || {
-            let repo = tramiton_core::Repo::new(&path);
-            tramiton_core::provider::analyze(&repo)
-        })
-        .await
-        .map_err(|e| CoreError::Other(format!("tramiton detect task join error: {e}")))?
-        .map_err(|e| CoreError::Other(format!("tramiton analyze error: {e}")))?;
-        Ok(plan.map(|bp| detection_from_build_plan(&bp)))
-    }
-}
-
-/// Map tramiton's `BuildPlan` onto our minimal detection summary.
-fn detection_from_build_plan(bp: &tramiton_core::BuildPlan) -> FirmwareDetection {
-    FirmwareDetection {
-        provider: bp.provider.clone(),
-        confidence: bp.confidence.to_string(),
-        build_system: bp.build_system.label().to_string(),
-        framework: bp.framework.clone(),
-        target: FirmwareTarget {
-            board: bp.target.board.clone(),
-            mcu: bp.target.mcu.clone(),
-            arch: bp.target.arch.clone(),
-        },
-        gaps: bp.gaps.clone(),
-    }
-}
-
-/// Map a firmware detection to a target type. Framework/build-system signals
-/// distinguish RTOS from bare-metal from Yocto.
-pub fn detection_to_target_type(det: &FirmwareDetection) -> TargetType {
-    let framework = det.framework.as_deref().unwrap_or("").to_lowercase();
-    let build_system = det.build_system.to_lowercase();
-    let signal = format!("{framework} {build_system} {}", det.provider.to_lowercase());
-
-    const RTOS: [&str; 6] = ["zephyr", "esp-idf", "freertos", "nuttx", "riot", "chibios"];
-    if signal.contains("bitbake") || signal.contains("yocto") || signal.contains("openembedded") {
-        TargetType::EmbeddedLinuxYocto
-    } else if RTOS.iter().any(|k| signal.contains(k)) {
-        TargetType::FirmwareRtos
-    } else {
-        TargetType::FirmwareBareMetal
-    }
-}
-
-/// Map tramiton's confidence label to a `[0,1]` score.
-fn confidence_score(label: &str) -> f32 {
-    match label.to_lowercase().as_str() {
-        "high" => 0.9,
-        "medium" => 0.6,
-        "low" => 0.3,
-        _ => 0.4,
-    }
-}
-
-/// Turn a firmware detection into a classifier verdict, carrying the MCU / board
-/// / build-system as facts.
-pub fn detection_to_verdict(det: &FirmwareDetection) -> ClassifierVerdict {
-    let target_type = detection_to_target_type(det);
-    let mut facts = vec![DetectedFact::new(
-        "build_system",
-        det.build_system.clone(),
-        "tramiton",
-    )];
-    if let Some(fw) = &det.framework {
-        facts.push(DetectedFact::new("framework", fw.clone(), "tramiton"));
-    }
-    if let Some(mcu) = &det.target.mcu {
-        facts.push(DetectedFact::new("mcu", mcu.clone(), "tramiton"));
-    }
-    if let Some(board) = &det.target.board {
-        facts.push(DetectedFact::new("board", board.clone(), "tramiton"));
-    }
-    if let Some(arch) = &det.target.arch {
-        facts.push(DetectedFact::new("arch", arch.clone(), "tramiton"));
-    }
-    ClassifierVerdict {
-        target_type,
-        confidence: confidence_score(&det.confidence),
-        facts,
-        rationale: format!(
-            "tramiton detected build system '{}'{}",
-            det.build_system,
-            det.framework
-                .as_ref()
-                .map(|f| format!(" (framework {f})"))
-                .unwrap_or_default()
-        ),
-    }
-}
-
-/// A deterministic [`FirmwareDetector`] for tests — returns a preset detection.
-pub struct MockFirmwareDetector {
-    /// The detection to return (or `None` for "no detection").
-    pub detection: Option<FirmwareDetection>,
-}
-
-impl FirmwareDetector for MockFirmwareDetector {
-    async fn detect(&self, _path: &Path) -> Result<Option<FirmwareDetection>, CoreError> {
-        Ok(self.detection.clone())
-    }
-}
-
-#[cfg(test)]
-#[allow(clippy::expect_used, clippy::unwrap_used)]
-mod tests {
-    use super::*;
-
-    fn detection(build_system: &str, framework: Option<&str>) -> FirmwareDetection {
-        FirmwareDetection {
-            provider: build_system.to_string(),
-            confidence: "high".to_string(),
-            build_system: build_system.to_string(),
-            framework: framework.map(|s| s.to_string()),
-            target: FirmwareTarget {
-                mcu: Some("stm32f429".to_string()),
-                ..Default::default()
-            },
-            gaps: Vec::new(),
-        }
-    }
-
-    #[test]
-    fn zephyr_maps_to_rtos() {
-        assert_eq!(
-            detection_to_target_type(&detection("zephyr", Some("zephyr"))),
-            TargetType::FirmwareRtos
-        );
-    }
-
-    #[test]
-    fn bare_cmake_maps_to_bare_metal() {
-        assert_eq!(
-            detection_to_target_type(&detection("cmake", Some("bare-metal"))),
-            TargetType::FirmwareBareMetal
-        );
-    }
-
-    #[test]
-    fn bitbake_maps_to_yocto() {
-        assert_eq!(
-            detection_to_target_type(&detection("bitbake", None)),
-            TargetType::EmbeddedLinuxYocto
-        );
-    }
-
-    #[test]
-    fn verdict_carries_mcu_fact_and_confidence() {
-        let v = detection_to_verdict(&detection("esp-idf", Some("esp-idf")));
-        assert_eq!(v.target_type, TargetType::FirmwareRtos);
-        assert!((v.confidence - 0.9).abs() < f32::EPSILON);
-        assert!(v
-            .facts
-            .iter()
-            .any(|f| f.key == "mcu" && f.value == "stm32f429"));
-    }
-}
@@ -1,357 +0,0 @@
-//! Heuristic target-type classification from artifact kinds and source markers.
-//!
-//! Complements the tramiton firmware detector: this handles web / backend /
-//! mobile / desktop / PLC by sniffing manifest files and file extensions in the
-//! ingested code trees, plus strong priors from the artifact kinds themselves
-//! (a PLC-project artifact is a PLC target; an `.ipa` is an iOS app).
-
-use std::collections::HashSet;
-use std::fs;
-use std::path::Path;
-
-use compliance_core::error::CoreError;
-use compliance_core::models::{ArtifactKind, DetectedFact, TargetType};
-use compliance_core::traits::{ClassificationInput, ClassifierVerdict, TargetClassifier};
-
-/// Max directory depth scanned for marker files.
-const SCAN_DEPTH: usize = 2;
-
-/// Markers collected from a code tree.
-#[derive(Default)]
-struct Markers {
-    files: HashSet<String>,
-    dirs: HashSet<String>,
-    exts: HashSet<String>,
-}
-
-impl Markers {
-    fn has_file(&self, name: &str) -> bool {
-        self.files.contains(name)
-    }
-    fn has_ext(&self, ext: &str) -> bool {
-        self.exts.contains(ext)
-    }
-    fn any_dir_ends_with(&self, suffix: &str) -> bool {
-        self.dirs.iter().any(|d| d.ends_with(suffix))
-    }
-}
-
-/// Recursively collect marker file/dir/extension names up to [`SCAN_DEPTH`].
-fn collect_markers(root: &Path) -> Markers {
-    let mut m = Markers::default();
-    scan_dir(root, 0, &mut m);
-    m
-}
-
-fn scan_dir(dir: &Path, depth: usize, m: &mut Markers) {
-    let Ok(entries) = fs::read_dir(dir) else {
-        return;
-    };
-    for entry in entries.flatten() {
-        let path = entry.path();
-        let name = entry.file_name().to_string_lossy().to_lowercase();
-        if path.is_dir() {
-            m.dirs.insert(name);
-            if depth < SCAN_DEPTH {
-                scan_dir(&path, depth + 1, m);
-            }
-        } else {
-            if let Some(ext) = path.extension() {
-                m.exts.insert(ext.to_string_lossy().to_lowercase());
-            }
-            m.files.insert(name);
-        }
-    }
-}
-
-/// Whether a `package.json` at `root` looks like a front-end app.
-fn package_json_is_frontend(root: &Path) -> bool {
-    let Ok(content) = fs::read_to_string(root.join("package.json")) else {
-        return false;
-    };
-    let c = content.to_lowercase();
-    ["react", "next", "vue", "@angular", "svelte", "vite"]
-        .iter()
-        .any(|f| c.contains(f))
-}
-
-/// The heuristic classifier: artifact-kind priors + source-tree markers.
-pub struct HeuristicClassifier;
-
-impl HeuristicClassifier {
-    /// Verdicts from the artifact kinds alone (no filesystem needed).
-    fn kind_priors(&self, input: &ClassificationInput<'_>) -> Vec<ClassifierVerdict> {
-        let mut out = Vec::new();
-        for a in input.artifacts {
-            let lower = a.source_ref.to_lowercase();
-            match a.kind {
-                ArtifactKind::PlcProject => out.push(verdict(
-                    TargetType::PlcSps,
-                    0.85,
-                    "PLC project artifact",
-                    vec![],
-                )),
-                ArtifactKind::MobilePackage => {
-                    let (tt, why) = if lower.ends_with(".ipa") {
-                        (TargetType::IosApp, "iOS package (.ipa)")
-                    } else {
-                        (TargetType::AndroidApp, "Android package (.apk/.aab)")
-                    };
-                    out.push(verdict(tt, 0.85, why, vec![]));
-                }
-                ArtifactKind::ContainerImage => out.push(verdict(
-                    TargetType::BackendService,
-                    0.4,
-                    "container image",
-                    vec![],
-                )),
-                ArtifactKind::FirmwareImage => out.push(verdict(
-                    TargetType::FirmwareBareMetal,
-                    0.35,
-                    "firmware image (pending tramiton detection)",
-                    vec![],
-                )),
-                ArtifactKind::LiveUrl if input.artifacts.len() == 1 => {
-                    out.push(verdict(TargetType::WebApp, 0.3, "live URL only", vec![]))
-                }
-                _ => {}
-            }
-        }
-        out
-    }
-
-    /// Verdicts from scanning the ingested code trees for manifest markers.
-    fn source_verdicts(&self, input: &ClassificationInput<'_>) -> Vec<ClassifierVerdict> {
-        let mut out = Vec::new();
-        for a in input.artifacts {
-            if !matches!(a.kind, ArtifactKind::GitRepo | ArtifactKind::SourceArchive) {
-                continue;
-            }
-            let Some(path) = input.working_paths.get(&a.id) else {
-                continue;
-            };
-            let m = collect_markers(path);
-
-            // Mobile (checked first — strongest signal).
-            if m.has_file("androidmanifest.xml") || m.has_ext("apk") || m.has_ext("aab") {
-                out.push(verdict(
-                    TargetType::AndroidApp,
-                    0.8,
-                    "Android manifest / gradle",
-                    facts_lang("kotlin/java"),
-                ));
-            }
-            if m.any_dir_ends_with(".xcodeproj")
-                || m.has_file("info.plist")
-                || m.has_file("podfile")
-                || m.has_ext("ipa")
-            {
-                out.push(verdict(
-                    TargetType::IosApp,
-                    0.8,
-                    "Xcode project / Info.plist",
-                    facts_lang("swift/objc"),
-                ));
-            }
-            // Desktop.
-            if m.has_ext("sln")
-                || m.has_ext("csproj")
-                || m.has_ext("vcxproj")
-                || m.has_ext("desktop")
-            {
-                out.push(verdict(
-                    TargetType::DesktopApp,
-                    0.7,
-                    "desktop project files",
-                    facts_lang("dotnet/native"),
-                ));
-            }
-            // PLC.
-            if m.has_ext("st") {
-                out.push(verdict(
-                    TargetType::PlcSps,
-                    0.8,
-                    "Structured Text sources",
-                    facts_lang("iec-61131-3"),
-                ));
-            }
-            // Web vs backend from package.json.
-            if m.has_file("package.json") {
-                if package_json_is_frontend(path) {
-                    out.push(verdict(
-                        TargetType::WebApp,
-                        0.65,
-                        "package.json with a front-end framework",
-                        facts_lang("javascript"),
-                    ));
-                } else {
-                    out.push(verdict(
-                        TargetType::BackendService,
-                        0.55,
-                        "package.json (no front-end framework)",
-                        facts_lang("javascript"),
-                    ));
-                }
-            }
-            // Backend languages.
-            for (file, lang) in [
-                ("cargo.toml", "rust"),
-                ("go.mod", "go"),
-                ("pom.xml", "java"),
-                ("requirements.txt", "python"),
-                ("pyproject.toml", "python"),
-            ] {
-                if m.has_file(file) {
-                    out.push(verdict(
-                        TargetType::BackendService,
-                        0.6,
-                        "backend build manifest",
-                        facts_lang(lang),
-                    ));
-                }
-            }
-            // Container-only.
-            if m.has_file("dockerfile") && out.is_empty() {
-                out.push(verdict(
-                    TargetType::BackendService,
-                    0.4,
-                    "Dockerfile",
-                    facts_lang("container"),
-                ));
-            }
-        }
-        out
-    }
-}
-
-impl TargetClassifier for HeuristicClassifier {
-    fn name(&self) -> &str {
-        "heuristic"
-    }
-
-    async fn classify(
-        &self,
-        input: &ClassificationInput<'_>,
-    ) -> Result<Vec<ClassifierVerdict>, CoreError> {
-        let mut out = self.kind_priors(input);
-        out.extend(self.source_verdicts(input));
-        Ok(out)
-    }
-}
-
-fn verdict(
-    target_type: TargetType,
-    confidence: f32,
-    rationale: &str,
-    facts: Vec<DetectedFact>,
-) -> ClassifierVerdict {
-    ClassifierVerdict {
-        target_type,
-        confidence,
-        facts,
-        rationale: rationale.to_string(),
-    }
-}
-
-fn facts_lang(lang: &str) -> Vec<DetectedFact> {
-    vec![DetectedFact::new("language", lang, "heuristic")]
-}
-
-#[cfg(test)]
-#[allow(clippy::expect_used, clippy::unwrap_used)]
-mod tests {
-    use super::*;
-    use compliance_core::models::Artifact;
-    use std::collections::HashMap;
-    use std::path::PathBuf;
-
-    struct Scratch(PathBuf);
-    impl Scratch {
-        fn new() -> Self {
-            let p = std::env::temp_dir().join(format!("cs-classify-{}", uuid::Uuid::new_v4()));
-            fs::create_dir_all(&p).expect("mkdir");
-            Self(p)
-        }
-    }
-    impl Drop for Scratch {
-        fn drop(&mut self) {
-            let _ = fs::remove_dir_all(&self.0);
-        }
-    }
-
-    async fn classify_tree(setup: impl FnOnce(&Path)) -> Vec<ClassifierVerdict> {
-        let scratch = Scratch::new();
-        setup(&scratch.0);
-        let artifact = Artifact::git_repo("https://git/x", "main");
-        let mut wp = HashMap::new();
-        wp.insert(artifact.id.clone(), scratch.0.clone());
-        let artifacts = vec![artifact];
-        let input = ClassificationInput {
-            artifacts: &artifacts,
-            working_paths: &wp,
-            description: None,
-        };
-        HeuristicClassifier
-            .classify(&input)
-            .await
-            .expect("classify")
-    }
-
-    #[tokio::test]
-    async fn frontend_package_json_is_webapp() {
-        let v = classify_tree(|root| {
-            fs::write(
-                root.join("package.json"),
-                r#"{"dependencies":{"react":"18"}}"#,
-            )
-            .unwrap();
-        })
-        .await;
-        assert!(v.iter().any(|x| x.target_type == TargetType::WebApp));
-    }
-
-    #[tokio::test]
-    async fn cargo_toml_is_backend() {
-        let v = classify_tree(|root| {
-            fs::write(root.join("Cargo.toml"), "[package]\nname='x'").unwrap();
-        })
-        .await;
-        assert!(v
-            .iter()
-            .any(|x| x.target_type == TargetType::BackendService));
-    }
-
-    #[tokio::test]
-    async fn android_manifest_is_android() {
-        let v = classify_tree(|root| {
-            fs::write(root.join("AndroidManifest.xml"), "<manifest/>").unwrap();
-        })
-        .await;
-        assert!(v.iter().any(|x| x.target_type == TargetType::AndroidApp));
-    }
-
-    #[tokio::test]
-    async fn structured_text_is_plc() {
-        let v = classify_tree(|root| {
-            fs::write(root.join("main.st"), "PROGRAM main END_PROGRAM").unwrap();
-        })
-        .await;
-        assert!(v.iter().any(|x| x.target_type == TargetType::PlcSps));
-    }
-
-    #[tokio::test]
-    async fn ipa_artifact_prior_is_ios() {
-        let artifacts = vec![Artifact::mobile_package("app.ipa")];
-        let wp = HashMap::new();
-        let input = ClassificationInput {
-            artifacts: &artifacts,
-            working_paths: &wp,
-            description: None,
-        };
-        let v = HeuristicClassifier
-            .classify(&input)
-            .await
-            .expect("classify");
-        assert!(v.iter().any(|x| x.target_type == TargetType::IosApp));
-    }
-}
@@ -1,226 +0,0 @@
-//! Target classification.
-//!
-//! Runs the classifier registry over a target's artifacts and their ingested
-//! working paths, then merges and ranks the verdicts into a [`Classification`].
-//! The registry is the heuristic classifier (artifact kinds + source markers)
-//! plus the tramiton firmware detector (behind a [`FirmwareDetector`] port).
-
-mod firmware;
-mod language;
-
-pub use firmware::{
-    FirmwareDetection, FirmwareDetector, FirmwareTarget, MockFirmwareDetector, TramitonNative,
-};
-pub use language::HeuristicClassifier;
-
-use std::collections::HashMap;
-use std::path::PathBuf;
-
-use compliance_core::error::CoreError;
-use compliance_core::models::{
-    ArtifactKind, Classification, DetectedFact, OnboardedTarget, TargetType, TargetTypeCandidate,
-};
-use compliance_core::traits::{ClassificationInput, ClassifierVerdict, TargetClassifier};
-
-use firmware::detection_to_verdict;
-
-/// Classify a target from its artifacts and their ingested working paths, using
-/// the heuristic classifier plus the tramiton firmware detector. Verdicts are
-/// merged (max confidence per target type) and ranked into a [`Classification`].
-pub async fn classify_target<D: FirmwareDetector>(
-    target: &OnboardedTarget,
-    working_paths: &HashMap<String, PathBuf>,
-    firmware_detector: &D,
-) -> Result<Classification, CoreError> {
-    let input = ClassificationInput {
-        artifacts: &target.artifacts,
-        working_paths,
-        description: target.description.as_deref(),
-    };
-
-    let mut verdicts = Vec::new();
-    let mut detected_by = Vec::new();
-
-    let heuristic = HeuristicClassifier.classify(&input).await?;
-    if !heuristic.is_empty() {
-        detected_by.push("heuristic".to_string());
-    }
-    verdicts.extend(heuristic);
-
-    // Tramiton firmware detection over firmware / code working paths.
-    let mut tramiton_used = false;
-    for artifact in &target.artifacts {
-        if !matches!(
-            artifact.kind,
-            ArtifactKind::FirmwareImage | ArtifactKind::GitRepo | ArtifactKind::SourceArchive
-        ) {
-            continue;
-        }
-        let Some(path) = working_paths.get(&artifact.id) else {
-            continue;
-        };
-        if let Some(detection) = firmware_detector.detect(path).await? {
-            verdicts.push(detection_to_verdict(&detection));
-            tramiton_used = true;
-        }
-    }
-    if tramiton_used {
-        detected_by.push("tramiton".to_string());
-    }
-
-    Ok(rank(verdicts, detected_by, target.target_type))
-}
-
-/// Merge verdicts by target type (keeping the max confidence and its rationale),
-/// dedupe facts, rank by descending confidence, and assemble a [`Classification`].
-/// Falls back to the declared type when no verdict is produced.
-fn rank(
-    verdicts: Vec<ClassifierVerdict>,
-    detected_by: Vec<String>,
-    fallback: TargetType,
-) -> Classification {
-    let mut best: HashMap<TargetType, (f32, String)> = HashMap::new();
-    let mut facts: Vec<DetectedFact> = Vec::new();
-    for verdict in verdicts {
-        for fact in verdict.facts {
-            if !facts
-                .iter()
-                .any(|e| e.key == fact.key && e.value == fact.value)
-            {
-                facts.push(fact);
-            }
-        }
-        let entry = best
-            .entry(verdict.target_type)
-            .or_insert((0.0, String::new()));
-        if verdict.confidence > entry.0 {
-            *entry = (verdict.confidence, verdict.rationale);
-        }
-    }
-
-    let mut candidates: Vec<TargetTypeCandidate> = best
-        .into_iter()
-        .map(
-            |(target_type, (confidence, rationale))| TargetTypeCandidate {
-                target_type,
-                confidence,
-                rationale,
-            },
-        )
-        .collect();
-    // Descending confidence; ties broken by type name for deterministic ordering.
-    candidates.sort_by(|a, b| {
-        b.confidence
-            .partial_cmp(&a.confidence)
-            .unwrap_or(std::cmp::Ordering::Equal)
-            .then_with(|| a.target_type.to_string().cmp(&b.target_type.to_string()))
-    });
-
-    let suggested = candidates
-        .first()
-        .map(|c| c.target_type)
-        .unwrap_or(fallback);
-
-    Classification {
-        suggested,
-        candidates,
-        facts,
-        detected_by,
-        detected_at: chrono::Utc::now(),
-        confirmed: false,
-    }
-}
-
-#[cfg(test)]
-#[allow(clippy::expect_used, clippy::unwrap_used)]
-mod tests {
-    use super::*;
-    use compliance_core::models::Artifact;
-    use std::fs;
-    use std::path::Path;
-
-    struct Scratch(PathBuf);
-    impl Scratch {
-        fn new() -> Self {
-            let p = std::env::temp_dir().join(format!("cs-classify-mod-{}", uuid::Uuid::new_v4()));
-            fs::create_dir_all(&p).expect("mkdir");
-            Self(p)
-        }
-    }
-    impl Drop for Scratch {
-        fn drop(&mut self) {
-            let _ = fs::remove_dir_all(&self.0);
-        }
-    }
-
-    fn no_firmware() -> MockFirmwareDetector {
-        MockFirmwareDetector { detection: None }
-    }
-
-    #[tokio::test]
-    async fn backend_repo_classifies_as_backend() {
-        let scratch = Scratch::new();
-        fs::write(scratch.0.join("go.mod"), "module x").unwrap();
-
-        let artifact = Artifact::git_repo("https://git/x", "main");
-        let mut wp = HashMap::new();
-        wp.insert(artifact.id.clone(), scratch.0.clone());
-        let mut target = OnboardedTarget::new("x".to_string(), TargetType::WebApp);
-        target.artifacts.push(artifact);
-
-        let c = classify_target(&target, &wp, &no_firmware())
-            .await
-            .expect("classify");
-        assert_eq!(c.suggested, TargetType::BackendService);
-        assert!(c.detected_by.contains(&"heuristic".to_string()));
-        assert!(!c.confirmed);
-    }
-
-    #[tokio::test]
-    async fn firmware_detector_verdict_ranks_top() {
-        let scratch = Scratch::new();
-        fs::write(scratch.0.join("fw.bin"), b"x").unwrap();
-
-        let artifact =
-            Artifact::firmware_image(scratch.0.join("fw.bin").to_string_lossy().to_string());
-        let mut wp = HashMap::new();
-        wp.insert(artifact.id.clone(), scratch.0.clone());
-        let mut target = OnboardedTarget::new("fw".to_string(), TargetType::FirmwareBareMetal);
-        target.artifacts.push(artifact);
-
-        let detector = MockFirmwareDetector {
-            detection: Some(FirmwareDetection {
-                provider: "zephyr".to_string(),
-                confidence: "high".to_string(),
-                build_system: "zephyr".to_string(),
-                framework: Some("zephyr".to_string()),
-                target: FirmwareTarget {
-                    mcu: Some("nrf52840".to_string()),
-                    ..Default::default()
-                },
-                gaps: vec![],
-            }),
-        };
-
-        let c = classify_target(&target, &wp, &detector)
-            .await
-            .expect("classify");
-        // tramiton's high-confidence RTOS verdict beats the weak firmware prior.
-        assert_eq!(c.suggested, TargetType::FirmwareRtos);
-        assert!(c.detected_by.contains(&"tramiton".to_string()));
-        assert!(c.facts.iter().any(|f| f.key == "mcu"));
-    }
-
-    #[tokio::test]
-    async fn no_signal_falls_back_to_declared_type() {
-        let scratch = Scratch::new();
-        let _ = Path::new(&scratch.0);
-        let target = OnboardedTarget::new("empty".to_string(), TargetType::DesktopApp);
-        let wp = HashMap::new();
-        let c = classify_target(&target, &wp, &no_firmware())
-            .await
-            .expect("classify");
-        assert_eq!(c.suggested, TargetType::DesktopApp);
-        assert!(c.candidates.is_empty());
-    }
-}
@@ -1,4 +1,3 @@
-use compliance_core::config::{BreakpilotConfig, PlcRuntimeConfig};
 use compliance_core::AgentConfig;
 use secrecy::SecretString;

@@ -43,11 +42,9 @@ pub fn load_config() -> Result<AgentConfig, AgentError> {
            .unwrap_or(3001),
        scan_schedule: env_var_opt("SCAN_SCHEDULE").unwrap_or_else(|| "0 0 */6 * * *".to_string()),
        cve_monitor_schedule: env_var_opt("CVE_MONITOR_SCHEDULE")
-            .unwrap_or_else(|| "0 0 * * * *".to_string()),
+            .unwrap_or_else(|| "0 0 0 * * *".to_string()),
        git_clone_base_path: env_var_opt("GIT_CLONE_BASE_PATH")
            .unwrap_or_else(|| "/tmp/compliance-scanner/repos".to_string()),
-        artifact_store_base_path: env_var_opt("ARTIFACT_STORE_BASE_PATH")
-            .unwrap_or_else(|| "/data/compliance-scanner/artifacts".to_string()),
        ssh_key_path: env_var_opt("SSH_KEY_PATH")
            .unwrap_or_else(|| "/data/compliance-scanner/ssh/id_ed25519".to_string()),
        keycloak_url: env_var_opt("KEYCLOAK_URL"),
@@ -62,49 +59,5 @@ pub fn load_config() -> Result<AgentConfig, AgentError> {
            .unwrap_or(true),
        pentest_imap_username: env_var_opt("PENTEST_IMAP_USERNAME"),
        pentest_imap_password: env_secret_opt("PENTEST_IMAP_PASSWORD"),
-        admin_api_token: env_secret_opt("ADMIN_API_TOKEN"),
-        tenant_registry_url: env_var_opt("TENANT_REGISTRY_URL"),
-        plc_runtime: load_plc_runtime_config(),
-        werkbank_runner_token: env_secret_opt("WERKBANK_RUNNER_TOKEN"),
-        breakpilot: load_breakpilot_config(),
    })
 }
-
-/// Build the ephemeral soft-PLC provisioning config from the environment,
-/// falling back to [`PlcRuntimeConfig::default`] for any unset knob. Disabled
-/// unless `PLC_RUNTIME_ENABLED` is truthy — it requires Docker access.
-fn load_plc_runtime_config() -> PlcRuntimeConfig {
-    let d = PlcRuntimeConfig::default();
-    PlcRuntimeConfig {
-        enabled: env_var_opt("PLC_RUNTIME_ENABLED")
-            .map(|v| v == "1" || v.eq_ignore_ascii_case("true"))
-            .unwrap_or(d.enabled),
-        image: env_var_opt("PLC_RUNTIME_IMAGE").unwrap_or(d.image),
-        network: env_var_opt("PLC_RUNTIME_NETWORK").unwrap_or(d.network),
-        memory: env_var_opt("PLC_RUNTIME_MEMORY").unwrap_or(d.memory),
-        cpus: env_var_opt("PLC_RUNTIME_CPUS").unwrap_or(d.cpus),
-        max_lifetime_secs: env_var_opt("PLC_RUNTIME_MAX_LIFETIME_SECS")
-            .and_then(|v| v.parse().ok())
-            .unwrap_or(d.max_lifetime_secs),
-        openplc_user: env_var_opt("PLC_RUNTIME_OPENPLC_USER").unwrap_or(d.openplc_user),
-        openplc_password: env_secret_opt("PLC_RUNTIME_OPENPLC_PASSWORD")
-            .unwrap_or(d.openplc_password),
-    }
-}
-
-/// Assemble the breakpilot OSCAL-catalog source from env, defaulting the snapshot
-/// directory. A missing `BREAKPILOT_BASE_URL` leaves the controls provider off.
-fn load_breakpilot_config() -> BreakpilotConfig {
-    let d = BreakpilotConfig::default();
-    BreakpilotConfig {
-        base_url: env_var_opt("BREAKPILOT_BASE_URL"),
-        token: env_secret_opt("BREAKPILOT_TOKEN"),
-        snapshot_dir: env_var_opt("BREAKPILOT_SNAPSHOT_DIR").unwrap_or(d.snapshot_dir),
-        semantic_mapping: env_var_opt("BREAKPILOT_SEMANTIC_MAPPING")
-            .map(|v| v == "1" || v.eq_ignore_ascii_case("true"))
-            .unwrap_or(d.semantic_mapping),
-        grounded_control_checks: env_var_opt("BREAKPILOT_GROUNDED_CHECKS")
-            .map(|v| v == "1" || v.eq_ignore_ascii_case("true"))
-            .unwrap_or(d.grounded_control_checks),
-    }
-}
@@ -1,114 +0,0 @@
-//! The grounded control checker: judge each candidate region for a control, then
-//! keep only the verdicts that survive the grounding gate.
-//!
-//! Generic over [`ControlJudge`] so tests drive it with a deterministic stub —
-//! the whole recognize → ground path is then exercised without an LLM. With the
-//! real judge, determinism comes from temperature 0 plus the gate.
-
-use compliance_core::control_check::{ground, CandidateRegion, ControlCheckSpec};
-use compliance_core::models::Finding;
-
-use super::judge::ControlJudge;
-
-/// Runs a [`ControlJudge`] over candidate regions and grounds the results.
-pub struct GroundedControlChecker<J> {
-    judge: J,
-}
-
-impl<J: ControlJudge> GroundedControlChecker<J> {
-    pub fn new(judge: J) -> Self {
-        Self { judge }
-    }
-
-    /// Judge every candidate region for `spec` and return the grounded findings.
-    /// A verdict that doesn't quote real code in its region is dropped by
-    /// [`ground`], so nothing fabricated reaches the caller.
-    pub async fn check(
-        &self,
-        spec: &ControlCheckSpec,
-        regions: &[CandidateRegion],
-        repo_id: &str,
-    ) -> Vec<Finding> {
-        let mut findings = Vec::new();
-        for region in regions {
-            let verdict = self.judge.judge(spec, region).await;
-            if let Some(finding) = ground(spec, region, &verdict, repo_id) {
-                findings.push(finding);
-            }
-        }
-        findings
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use compliance_core::control_check::LlmVerdict;
-    use compliance_core::models::finding::Severity;
-
-    /// Deterministic stub: returns a fixed verdict for every region, so the
-    /// recognize → ground composition is tested without an LLM.
-    struct StubJudge {
-        verdict: LlmVerdict,
-    }
-
-    impl ControlJudge for StubJudge {
-        async fn judge(&self, _spec: &ControlCheckSpec, _region: &CandidateRegion) -> LlmVerdict {
-            self.verdict.clone()
-        }
-    }
-
-    fn spec() -> ControlCheckSpec {
-        ControlCheckSpec {
-            control_id: "cra-ai-8".into(),
-            title: "No default passwords".into(),
-            requirement: "No default credentials".into(),
-            default_cwe: Some("CWE-798".into()),
-            severity: Severity::High,
-        }
-    }
-
-    fn region(content: &str) -> CandidateRegion {
-        CandidateRegion {
-            file: "src/auth.py".into(),
-            start_line: 1,
-            content: content.into(),
-        }
-    }
-
-    #[tokio::test]
-    async fn keeps_grounded_and_drops_ungrounded() {
-        let checker = GroundedControlChecker::new(StubJudge {
-            verdict: LlmVerdict {
-                violates: true,
-                snippet: "PASSWORD = \"admin\"".into(),
-                cwe: None,
-                confidence: 0.9,
-            },
-        });
-        let regions = vec![
-            region("x = 1\nPASSWORD = \"admin\"\n"), // quotes real code → grounded
-            region("totally unrelated code\n"),      // snippet absent → dropped
-        ];
-        let findings = checker.check(&spec(), &regions, "repo").await;
-        assert_eq!(findings.len(), 1);
-        assert_eq!(findings[0].control_refs, vec!["cra-ai-8".to_string()]);
-        assert_eq!(findings[0].line_number, Some(2));
-    }
-
-    #[tokio::test]
-    async fn non_violation_yields_nothing() {
-        let checker = GroundedControlChecker::new(StubJudge {
-            verdict: LlmVerdict {
-                violates: false,
-                snippet: String::new(),
-                cwe: None,
-                confidence: 0.0,
-            },
-        });
-        let findings = checker
-            .check(&spec(), &[region("PASSWORD = \"admin\"\n")], "repo")
-            .await;
-        assert!(findings.is_empty());
-    }
-}
@@ -1,245 +0,0 @@
-//! In-memory embedding index over the control corpus, for region → control
-//! retrieval.
-//!
-//! At master-control scale (~13.6k) findings can't be mapped by CWE (the master
-//! controls carry none), so we map by *similarity*: embed each control's
-//! requirement text once, then for a code region pull the top-K nearest controls
-//! to hand to the grounded judge. This is the retrieval half of the semantic path.
-
-use std::path::Path;
-
-use serde::{Deserialize, Serialize};
-use sha2::{Digest, Sha256};
-
-use compliance_core::control_check::ControlCheckSpec;
-use compliance_core::error::CoreError;
-
-use crate::llm::LlmClient;
-
-/// A control spec paired with its requirement-text embedding.
-pub struct ControlIndex {
-    entries: Vec<(ControlCheckSpec, Vec<f64>)>,
-}
-
-/// On-disk form of the index: the corpus identity hash plus every spec+embedding.
-/// The hash lets a later scan reuse the embeddings only if the corpus is unchanged.
-#[derive(Serialize, Deserialize)]
-struct PersistedIndex {
-    corpus_hash: String,
-    entries: Vec<PersistedEntry>,
-}
-
-#[derive(Serialize, Deserialize)]
-struct PersistedEntry {
-    spec: ControlCheckSpec,
-    embedding: Vec<f64>,
-}
-
-/// Stable hash of the corpus identity (each control's id + requirement text, in
-/// order). Same catalog → same hash → the cached embeddings are reused instead of
-/// re-embedding the whole corpus.
-fn corpus_hash(specs: &[ControlCheckSpec]) -> String {
-    let mut hasher = Sha256::new();
-    for s in specs {
-        hasher.update(s.control_id.as_bytes());
-        hasher.update([0u8]);
-        hasher.update(s.requirement.as_bytes());
-        hasher.update([0u8]);
-    }
-    format!("{:x}", hasher.finalize())
-}
-
-impl ControlIndex {
-    /// Build directly from precomputed embeddings (used by tests + callers that
-    /// already embedded the corpus).
-    pub fn from_embeddings(entries: Vec<(ControlCheckSpec, Vec<f64>)>) -> Self {
-        Self { entries }
-    }
-
-    /// Load the index from `cache_path` if it still matches the current corpus,
-    /// otherwise embed the corpus and persist it there. This turns the per-scan
-    /// re-embed of the whole (~13.6k) master-control corpus into a one-time cost
-    /// that survives across scans; the cache self-invalidates when the catalog
-    /// changes (its [`corpus_hash`] no longer matches).
-    pub async fn load_or_build(
-        llm: &LlmClient,
-        specs: Vec<ControlCheckSpec>,
-        cache_path: &Path,
-    ) -> Result<Self, CoreError> {
-        let hash = corpus_hash(&specs);
-        if let Some(index) = Self::load_cache(cache_path, &hash).await {
-            tracing::debug!(
-                controls = index.len(),
-                "reusing cached control embedding index"
-            );
-            return Ok(index);
-        }
-        let index = Self::build(llm, specs).await?;
-        if let Err(e) = index.write_cache(cache_path, &hash).await {
-            tracing::warn!(error = %e, "failed to persist control embedding index");
-        }
-        Ok(index)
-    }
-
-    /// Read a persisted index, returning it only if its corpus hash matches.
-    async fn load_cache(path: &Path, hash: &str) -> Option<Self> {
-        let raw = tokio::fs::read(path).await.ok()?;
-        let persisted: PersistedIndex = serde_json::from_slice(&raw).ok()?;
-        if persisted.corpus_hash != hash {
-            return None;
-        }
-        Some(Self {
-            entries: persisted
-                .entries
-                .into_iter()
-                .map(|e| (e.spec, e.embedding))
-                .collect(),
-        })
-    }
-
-    /// Persist the index atomically (temp file + rename) keyed by corpus hash.
-    async fn write_cache(&self, path: &Path, hash: &str) -> Result<(), CoreError> {
-        if let Some(parent) = path.parent() {
-            tokio::fs::create_dir_all(parent).await?;
-        }
-        let persisted = PersistedIndex {
-            corpus_hash: hash.to_string(),
-            entries: self
-                .entries
-                .iter()
-                .map(|(spec, emb)| PersistedEntry {
-                    spec: spec.clone(),
-                    embedding: emb.clone(),
-                })
-                .collect(),
-        };
-        let raw = serde_json::to_vec(&persisted)?;
-        let tmp = path.with_extension("json.tmp");
-        tokio::fs::write(&tmp, &raw).await?;
-        tokio::fs::rename(&tmp, path).await?;
-        Ok(())
-    }
-
-    /// Build by embedding each control's requirement text.
-    pub async fn build(llm: &LlmClient, specs: Vec<ControlCheckSpec>) -> Result<Self, CoreError> {
-        if specs.is_empty() {
-            return Ok(Self {
-                entries: Vec::new(),
-            });
-        }
-        let texts: Vec<String> = specs.iter().map(|s| s.requirement.clone()).collect();
-        let embeddings = llm
-            .embed(texts)
-            .await
-            .map_err(|e| CoreError::Llm(e.to_string()))?;
-        Ok(Self {
-            entries: specs.into_iter().zip(embeddings).collect(),
-        })
-    }
-
-    pub fn len(&self) -> usize {
-        self.entries.len()
-    }
-
-    pub fn is_empty(&self) -> bool {
-        self.entries.is_empty()
-    }
-
-    /// The top-`k` control specs whose embedding is nearest (cosine) to `query`.
-    pub fn nearest(&self, query: &[f64], k: usize) -> Vec<ControlCheckSpec> {
-        let mut scored: Vec<(f64, &ControlCheckSpec)> = self
-            .entries
-            .iter()
-            .map(|(spec, emb)| (cosine(query, emb), spec))
-            .collect();
-        scored.sort_by(|a, b| b.0.total_cmp(&a.0));
-        scored.into_iter().take(k).map(|(_, s)| s.clone()).collect()
-    }
-}
-
-/// Cosine similarity; 0.0 for length-mismatched, empty, or zero vectors.
-fn cosine(a: &[f64], b: &[f64]) -> f64 {
-    if a.len() != b.len() || a.is_empty() {
-        return 0.0;
-    }
-    let dot: f64 = a.iter().zip(b).map(|(x, y)| x * y).sum();
-    let na: f64 = a.iter().map(|x| x * x).sum();
-    let nb: f64 = b.iter().map(|x| x * x).sum();
-    if na == 0.0 || nb == 0.0 {
-        return 0.0;
-    }
-    dot / (na.sqrt() * nb.sqrt())
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use compliance_core::models::finding::Severity;
-
-    fn spec(id: &str) -> ControlCheckSpec {
-        ControlCheckSpec {
-            control_id: id.into(),
-            title: id.into(),
-            requirement: id.into(),
-            default_cwe: None,
-            severity: Severity::Medium,
-        }
-    }
-
-    #[test]
-    fn nearest_ranks_by_cosine() {
-        let index = ControlIndex::from_embeddings(vec![
-            (spec("a"), vec![1.0, 0.0]),
-            (spec("b"), vec![0.0, 1.0]),
-            (spec("c"), vec![0.7, 0.7]),
-        ]);
-        let hits = index.nearest(&[0.9, 0.1], 2);
-        assert_eq!(hits.len(), 2);
-        assert_eq!(hits[0].control_id, "a"); // closest to [0.9,0.1]
-    }
-
-    #[test]
-    fn cosine_edges_are_zero() {
-        assert_eq!(cosine(&[1.0], &[1.0, 2.0]), 0.0); // length mismatch
-        assert_eq!(cosine(&[0.0, 0.0], &[1.0, 1.0]), 0.0); // zero vector
-        assert!((cosine(&[1.0, 0.0], &[1.0, 0.0]) - 1.0).abs() < 1e-9); // identical
-    }
-
-    #[test]
-    fn corpus_hash_is_stable_and_identity_sensitive() {
-        let a = corpus_hash(&[spec("x"), spec("y")]);
-        assert_eq!(a, corpus_hash(&[spec("x"), spec("y")])); // same corpus → same hash
-        assert_ne!(a, corpus_hash(&[spec("y"), spec("x")])); // reorder → different
-        assert_ne!(a, corpus_hash(&[spec("x")])); // fewer controls → different
-    }
-
-    #[tokio::test]
-    #[allow(clippy::unwrap_used)]
-    async fn cache_round_trips_and_misses_on_corpus_change() {
-        let dir = std::env::temp_dir().join(format!("cidx-{}", uuid::Uuid::new_v4()));
-        let path = dir.join("control-index.json");
-        let specs = [spec("a"), spec("b")];
-        let hash = corpus_hash(&specs);
-        let index = ControlIndex::from_embeddings(vec![
-            (spec("a"), vec![1.0, 0.0]),
-            (spec("b"), vec![0.0, 1.0]),
-        ]);
-        index.write_cache(&path, &hash).await.unwrap();
-
-        // matching corpus hash → hit
-        let loaded = ControlIndex::load_cache(&path, &hash).await.unwrap();
-        assert_eq!(loaded.len(), 2);
-        assert_eq!(loaded.nearest(&[0.9, 0.1], 1)[0].control_id, "a");
-        // corpus changed → miss (forces a rebuild)
-        assert!(ControlIndex::load_cache(&path, "differenthash")
-            .await
-            .is_none());
-        // absent file → miss, not an error
-        assert!(
-            ControlIndex::load_cache(dir.join("nope.json").as_path(), &hash)
-                .await
-                .is_none()
-        );
-        let _ = std::fs::remove_dir_all(&dir);
-    }
-}
@@ -1,167 +0,0 @@
-//! The "recognize" stage: judge whether a code region violates a control.
-//!
-//! Behind the [`ControlJudge`] trait so the grounded checker can be driven by a
-//! deterministic stub in tests. The real [`LlmControlJudge`] runs the model at
-//! temperature 0 with a closed prompt — it must quote the offending code VERBATIM,
-//! and everything it returns is then re-checked by the grounding gate
-//! ([`compliance_core::control_check::ground`]). The judge is allowed to be
-//! smart; it is never trusted.
-
-use std::sync::Arc;
-
-use serde::Deserialize;
-
-use compliance_core::control_check::{CandidateRegion, ControlCheckSpec, LlmVerdict};
-
-use crate::llm::LlmClient;
-
-/// Prompt/logic version — part of the verdict cache key, bump on any change here.
-pub const PROMPT_VERSION: &str = "control-judge-v1";
-
-const SYSTEM_PROMPT: &str = "You are a precise security & compliance code auditor. \
-You are given ONE compliance control (a requirement) and ONE code region. Decide \
-ONLY whether the code region VIOLATES the control. Rules: (1) Judge only the code \
-shown — never assume code that is not present. (2) If and only if it violates, copy \
-the EXACT offending code VERBATIM into `snippet`, character-for-character from the \
-region — do not paraphrase, reformat, or reconstruct it. (3) If it does not clearly \
-violate, set violates=false and leave snippet empty. (4) Prefer false over guessing. \
-Respond with STRICT JSON only, no prose: \
-{\"violates\": bool, \"snippet\": \"<verbatim code or empty>\", \"cwe\": \"CWE-NNN or null\", \"confidence\": 0.0-1.0}";
-
-/// Judges one (control, region). Async-in-trait so a stub can drive tests.
-#[allow(async_fn_in_trait)]
-pub trait ControlJudge: Send + Sync {
-    async fn judge(&self, spec: &ControlCheckSpec, region: &CandidateRegion) -> LlmVerdict;
-}
-
-/// The real judge: the LLM at temperature 0 with the closed, verbatim-snippet prompt.
-pub struct LlmControlJudge {
-    llm: Arc<LlmClient>,
-}
-
-impl LlmControlJudge {
-    pub fn new(llm: Arc<LlmClient>) -> Self {
-        Self { llm }
-    }
-}
-
-impl ControlJudge for LlmControlJudge {
-    async fn judge(&self, spec: &ControlCheckSpec, region: &CandidateRegion) -> LlmVerdict {
-        let user = build_user_prompt(spec, region);
-        match self.llm.chat(SYSTEM_PROMPT, &user, Some(0.0)).await {
-            Ok(response) => parse_verdict(&response),
-            Err(e) => {
-                // Fail closed: a transient model error yields no finding, never a
-                // fabricated one.
-                tracing::warn!(control = %spec.control_id, error = %e, "control judge call failed");
-                no_violation()
-            }
-        }
-    }
-}
-
-fn build_user_prompt(spec: &ControlCheckSpec, region: &CandidateRegion) -> String {
-    format!(
-        "CONTROL {id} — {title}\nRequirement: {req}\n\nCODE ({file}, first line = {line}):\n```\n{code}\n```\n\nReturn the JSON verdict.",
-        id = spec.control_id,
-        title = spec.title,
-        req = spec.requirement,
-        file = region.file,
-        line = region.start_line,
-        code = region.content,
-    )
-}
-
-#[derive(Debug, Default, Deserialize)]
-struct RawVerdict {
-    #[serde(default)]
-    violates: bool,
-    #[serde(default)]
-    snippet: String,
-    #[serde(default)]
-    cwe: Option<String>,
-    #[serde(default)]
-    confidence: f64,
-}
-
-/// Parse the model's JSON verdict, tolerant of ```json fencing. Any parse failure
-/// degrades to a non-violation (never a fabricated finding).
-fn parse_verdict(response: &str) -> LlmVerdict {
-    let cleaned = response
-        .trim()
-        .trim_start_matches("```json")
-        .trim_start_matches("```")
-        .trim_end_matches("```")
-        .trim();
-    match serde_json::from_str::<RawVerdict>(cleaned) {
-        Ok(raw) => LlmVerdict {
-            violates: raw.violates,
-            snippet: raw.snippet,
-            cwe: raw.cwe.filter(|c| !c.trim().is_empty()),
-            confidence: raw.confidence,
-        },
-        Err(e) => {
-            tracing::debug!(error = %e, "failed to parse control verdict; treating as non-violation");
-            no_violation()
-        }
-    }
-}
-
-fn no_violation() -> LlmVerdict {
-    LlmVerdict {
-        violates: false,
-        snippet: String::new(),
-        cwe: None,
-        confidence: 0.0,
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use compliance_core::models::finding::Severity;
-
-    fn spec() -> ControlCheckSpec {
-        ControlCheckSpec {
-            control_id: "cra-ai-8".into(),
-            title: "No default passwords".into(),
-            requirement: "Products must not ship default credentials".into(),
-            default_cwe: Some("CWE-798".into()),
-            severity: Severity::High,
-        }
-    }
-
-    #[test]
-    fn parses_plain_and_fenced_json() {
-        let plain = r#"{"violates": true, "snippet": "PASSWORD = \"x\"", "cwe": "CWE-798", "confidence": 0.9}"#;
-        let v = parse_verdict(plain);
-        assert!(v.violates);
-        assert_eq!(v.snippet, "PASSWORD = \"x\"");
-        assert_eq!(v.cwe.as_deref(), Some("CWE-798"));
-
-        let fenced = "```json\n{\"violates\": false, \"snippet\": \"\", \"cwe\": null, \"confidence\": 0.1}\n```";
-        assert!(!parse_verdict(fenced).violates);
-    }
-
-    #[test]
-    fn garbage_and_empty_cwe_are_safe() {
-        assert!(!parse_verdict("not json at all").violates); // fail closed
-        let no_cwe =
-            parse_verdict(r#"{"violates": true, "snippet": "x", "cwe": "  ", "confidence": 0.5}"#);
-        assert!(no_cwe.cwe.is_none()); // blank CWE normalised away
-    }
-
-    #[test]
-    fn user_prompt_carries_control_and_code() {
-        let region = CandidateRegion {
-            file: "src/auth.py".into(),
-            start_line: 10,
-            content: "PASSWORD = \"admin\"".into(),
-        };
-        let p = build_user_prompt(&spec(), &region);
-        assert!(p.contains("cra-ai-8"));
-        assert!(p.contains("Products must not ship default credentials"));
-        assert!(p.contains("PASSWORD = \"admin\""));
-        assert!(p.contains("src/auth.py"));
-    }
-}
@@ -1,23 +0,0 @@
-//! Controls corpus providers.
-//!
-//! Implementations of [`compliance_core::traits::ControlsProvider`] that supply
-//! the control corpus the mapping engine assesses findings against. Currently:
-//! [`OscalControlsProvider`], which pulls breakpilot-compliance's OSCAL catalog
-//! and snapshots it locally.
-
-mod checker;
-mod index;
-mod judge;
-mod oscal_provider;
-mod scan_triage;
-mod semantic;
-mod surface;
-mod triage;
-
-pub use checker::GroundedControlChecker;
-pub use index::ControlIndex;
-pub use judge::{ControlJudge, LlmControlJudge, PROMPT_VERSION};
-pub use oscal_provider::OscalControlsProvider;
-pub use scan_triage::{grounded_surface_findings, semantic_stamp_findings, triage_repo_findings};
-pub use semantic::SemanticControlChecker;
-pub use triage::{ControlTriage, TriageOutcome};
@@ -1,229 +0,0 @@
-//! Pull + snapshot [`ControlsProvider`] backed by breakpilot-compliance's OSCAL
-//! catalog export.
-//!
-//! Fetches `GET {base}/api/compliance/v1/oscal/catalog?framework=<fw>`, snapshots
-//! the exact bytes to disk (so scans are deterministic and keep working offline /
-//! on-prem), and maps the catalog into the corpus controls the mapping engine
-//! consumes. The producer owns the catalog; we own the assessment — this is the
-//! ingest half of the loop.
-
-use std::path::PathBuf;
-
-use secrecy::{ExposeSecret, SecretString};
-
-use compliance_core::error::CoreError;
-use compliance_core::models::onboarding::ComplianceFramework;
-use compliance_core::models::oscal::OscalDocument;
-use compliance_core::traits::{Control, ControlQuery, ControlsProvider};
-
-/// A [`ControlsProvider`] that pulls the OSCAL catalog from breakpilot-compliance
-/// and snapshots it locally for deterministic / offline reuse.
-pub struct OscalControlsProvider {
-    http: reqwest::Client,
-    base_url: String,
-    token: Option<SecretString>,
-    snapshot_dir: PathBuf,
-}
-
-impl OscalControlsProvider {
-    /// Create a provider. `base_url` is the breakpilot-compliance root (e.g.
-    /// `http://backend-compliance:8002`); `snapshot_dir` is where catalog
-    /// snapshots are written so a later scan can reuse them without the network.
-    pub fn new(
-        http: reqwest::Client,
-        base_url: impl Into<String>,
-        token: Option<SecretString>,
-        snapshot_dir: impl Into<PathBuf>,
-    ) -> Self {
-        Self {
-            http,
-            base_url: base_url.into(),
-            token,
-            snapshot_dir: snapshot_dir.into(),
-        }
-    }
-
-    fn catalog_url(&self, framework: &str) -> String {
-        format!(
-            "{}/api/compliance/v1/oscal/catalog?framework={framework}",
-            self.base_url.trim_end_matches('/')
-        )
-    }
-
-    fn snapshot_path(&self, framework: &str) -> PathBuf {
-        self.snapshot_dir
-            .join(format!("oscal-catalog-{framework}.json"))
-    }
-
-    /// Fetch the raw catalog bytes for a framework token over HTTP.
-    async fn fetch_raw(&self, framework: &str) -> Result<Vec<u8>, CoreError> {
-        let mut req = self.http.get(self.catalog_url(framework));
-        if let Some(token) = &self.token {
-            req = req.bearer_auth(token.expose_secret());
-        }
-        let resp = req
-            .send()
-            .await
-            .map_err(|e| CoreError::Http(e.to_string()))?;
-        if !resp.status().is_success() {
-            return Err(CoreError::Http(format!(
-                "catalog fetch for {framework} returned HTTP {}",
-                resp.status()
-            )));
-        }
-        resp.bytes()
-            .await
-            .map(|b| b.to_vec())
-            .map_err(|e| CoreError::Http(e.to_string()))
-    }
-
-    /// Write a catalog snapshot atomically (temp file + rename).
-    async fn write_snapshot(&self, framework: &str, raw: &[u8]) -> Result<(), CoreError> {
-        tokio::fs::create_dir_all(&self.snapshot_dir).await?;
-        let path = self.snapshot_path(framework);
-        let tmp = path.with_extension("json.tmp");
-        tokio::fs::write(&tmp, raw).await?;
-        tokio::fs::rename(&tmp, &path).await?;
-        Ok(())
-    }
-
-    /// Read a previously written snapshot, if one exists.
-    async fn read_snapshot(&self, framework: &str) -> Result<Option<OscalDocument>, CoreError> {
-        match tokio::fs::read(self.snapshot_path(framework)).await {
-            Ok(raw) => Ok(Some(serde_json::from_slice(&raw)?)),
-            Err(e) if e.kind() == std::io::ErrorKind::NotFound => Ok(None),
-            Err(e) => Err(e.into()),
-        }
-    }
-
-    /// Load the catalog for a framework token: fetch fresh + snapshot the exact
-    /// bytes; on network failure, fall back to the last snapshot so scans run.
-    async fn load_token(&self, framework: &str) -> Result<OscalDocument, CoreError> {
-        match self.fetch_raw(framework).await {
-            Ok(raw) => {
-                let doc: OscalDocument = serde_json::from_slice(&raw)?;
-                if let Err(e) = self.write_snapshot(framework, &raw).await {
-                    tracing::warn!(framework, error = %e, "failed to write OSCAL snapshot");
-                }
-                Ok(doc)
-            }
-            Err(fetch_err) => match self.read_snapshot(framework).await? {
-                Some(doc) => {
-                    tracing::warn!(
-                        framework, error = %fetch_err,
-                        "OSCAL catalog fetch failed; falling back to snapshot"
-                    );
-                    Ok(doc)
-                }
-                None => Err(fetch_err),
-            },
-        }
-    }
-
-    /// Load the OSCAL catalog for a compliance framework.
-    pub async fn load(&self, framework: ComplianceFramework) -> Result<OscalDocument, CoreError> {
-        self.load_token(&framework.to_string()).await
-    }
-
-    /// Load the code-checkable master-controls catalog
-    /// (`?framework=master-controls`).
-    pub async fn load_master_controls(&self) -> Result<OscalDocument, CoreError> {
-        self.load_token("master-controls").await
-    }
-}
-
-/// Order controls whose title/text mention the query context first (stable), then
-/// truncate to the requested limit. Naive relevance — refined when the assessment
-/// layer lands.
-fn rank_and_truncate(mut controls: Vec<Control>, context: &str, limit: usize) -> Vec<Control> {
-    if !context.is_empty() {
-        let needle = context.to_lowercase();
-        controls.sort_by_key(|c| {
-            let hit =
-                c.title.to_lowercase().contains(&needle) || c.text.to_lowercase().contains(&needle);
-            u8::from(!hit)
-        });
-    }
-    controls.truncate(limit);
-    controls
-}
-
-impl ControlsProvider for OscalControlsProvider {
-    fn name(&self) -> &str {
-        "breakpilot-oscal"
-    }
-
-    async fn controls(&self, query: &ControlQuery<'_>) -> Result<Vec<Control>, CoreError> {
-        let mut out: Vec<Control> = Vec::new();
-        for &framework in query.frameworks {
-            match self.load(framework).await {
-                Ok(doc) => out.extend(doc.to_controls()),
-                Err(e) => {
-                    tracing::warn!(%framework, error = %e, "skipping framework: catalog unavailable")
-                }
-            }
-        }
-        Ok(rank_and_truncate(out, query.context, query.limit))
-    }
-}
-
-#[cfg(test)]
-#[allow(clippy::unwrap_used)]
-mod tests {
-    use super::*;
-
-    const MINI_CATALOG: &str = r#"{"catalog":{"uuid":"u","metadata":{"title":"T",
-      "version":"1.0.0","oscal-version":"1.1.2","props":[{"name":"framework","value":"cra"}]},
-      "groups":[{"id":"g","title":"G","controls":[{"id":"cra-ai-1","title":"MFA",
-      "props":[],"parts":[{"name":"statement","prose":"require mfa"}]}]}]}}"#;
-
-    fn provider(dir: &std::path::Path) -> OscalControlsProvider {
-        OscalControlsProvider::new(reqwest::Client::new(), "http://unused/", None, dir)
-    }
-
-    #[test]
-    fn builds_catalog_url_and_snapshot_path() {
-        let p = provider(std::path::Path::new("/snap"));
-        assert_eq!(
-            p.catalog_url("cra"),
-            "http://unused/api/compliance/v1/oscal/catalog?framework=cra"
-        );
-        assert_eq!(
-            p.snapshot_path("cra"),
-            std::path::Path::new("/snap/oscal-catalog-cra.json")
-        );
-    }
-
-    #[test]
-    fn ranks_context_hits_first_then_truncates() {
-        let mk = |id: &str, title: &str| Control {
-            id: id.into(),
-            framework: ComplianceFramework::Cra,
-            title: title.into(),
-            text: String::new(),
-            source: None,
-        };
-        let controls = vec![
-            mk("a", "logging policy"),
-            mk("b", "multi-factor auth"),
-            mk("c", "backup"),
-        ];
-        let ranked = rank_and_truncate(controls, "auth", 2);
-        assert_eq!(ranked.len(), 2);
-        assert_eq!(ranked[0].id, "b"); // the "auth" hit floats to the top
-    }
-
-    #[tokio::test]
-    async fn snapshot_round_trip_and_offline_fallback() {
-        let dir = std::env::temp_dir().join(format!("oscal-test-{}", uuid::Uuid::new_v4()));
-        let p = provider(&dir);
-        assert!(p.read_snapshot("cra").await.unwrap().is_none());
-        p.write_snapshot("cra", MINI_CATALOG.as_bytes())
-            .await
-            .unwrap();
-        let doc = p.read_snapshot("cra").await.unwrap().unwrap();
-        assert_eq!(doc.to_controls().len(), 1);
-        assert_eq!(doc.framework(), Some(ComplianceFramework::Cra));
-        let _ = std::fs::remove_dir_all(&dir);
-    }
-}
@@ -1,299 +0,0 @@
-//! Scan-pipeline integration for control triage.
-//!
-//! After the deterministic tools have produced findings, this stamps each finding
-//! with the compliance control(s) it's evidence for and marks control-level false
-//! positives — using the ingested OSCAL catalog for control text, the
-//! `control-map` LUT for the finding→control link, and the grounded LLM judge to
-//! confirm. Skipped entirely unless breakpilot is configured.
-
-use std::collections::HashMap;
-use std::path::Path;
-use std::sync::Arc;
-
-use compliance_core::control_check::{CandidateRegion, ControlCheckSpec};
-use compliance_core::models::finding::{Finding, FindingStatus, Severity};
-use compliance_core::models::onboarding::ComplianceFramework;
-use compliance_core::AgentConfig;
-use control_map::ControlMap;
-
-use super::surface;
-use super::{
-    ControlIndex, ControlTriage, GroundedControlChecker, LlmControlJudge, OscalControlsProvider,
-    SemanticControlChecker, TriageOutcome,
-};
-use crate::llm::LlmClient;
-
-/// Nearest master controls judged per code region in the semantic pass.
-const SEMANTIC_TOP_K: usize = 5;
-
-/// Lines of context to read on each side of a finding's line.
-const REGION_WINDOW: usize = 6;
-
-/// Triage every finding in `findings` against the CRA control map: stamp
-/// `control_refs` on confirmed findings and flag control false positives. Returns
-/// the number of findings tagged with at least one control.
-pub async fn triage_repo_findings(
-    config: &AgentConfig,
-    llm: Arc<LlmClient>,
-    repo_path: &Path,
-    findings: &mut [Finding],
-) -> usize {
-    let Some(base_url) = config.breakpilot.base_url.clone() else {
-        return 0; // control triage is opt-in via BREAKPILOT_BASE_URL
-    };
-    let provider = OscalControlsProvider::new(
-        reqwest::Client::new(),
-        base_url,
-        config.breakpilot.token.clone(),
-        &config.breakpilot.snapshot_dir,
-    );
-    let specs = build_specs(&provider).await;
-    if specs.is_empty() {
-        return 0;
-    }
-    let map = match ControlMap::cra() {
-        Ok(m) => m,
-        Err(e) => {
-            tracing::warn!(error = %e, "control map failed to load; skipping control triage");
-            return 0;
-        }
-    };
-    let triage = ControlTriage::new(LlmControlJudge::new(llm), map, specs);
-
-    let mut tagged = 0;
-    for finding in findings.iter_mut() {
-        let (Some(file), Some(line)) = (finding.file_path.clone(), finding.line_number) else {
-            continue;
-        };
-        let Some(region) = fetch_region(repo_path, &file, line) else {
-            continue;
-        };
-        match triage.triage(finding, &region).await {
-            TriageOutcome::Confirmed(controls) => {
-                finding.control_refs = controls;
-                tagged += 1;
-            }
-            TriageOutcome::FalsePositive => {
-                finding.status = FindingStatus::FalsePositive;
-                finding.triage_action = Some("control_false_positive".to_string());
-            }
-            TriageOutcome::Unmapped => {}
-        }
-    }
-    tagged
-}
-
-/// Build the control requirement specs (by id) from the ingested OSCAL catalog.
-async fn build_specs(provider: &OscalControlsProvider) -> HashMap<String, ControlCheckSpec> {
-    let mut specs = HashMap::new();
-    match provider.load(ComplianceFramework::Cra).await {
-        Ok(doc) => {
-            for control in doc.to_controls() {
-                specs.insert(
-                    control.id.clone(),
-                    ControlCheckSpec {
-                        control_id: control.id,
-                        title: control.title,
-                        requirement: control.text,
-                        default_cwe: None,
-                        severity: Severity::Medium,
-                    },
-                );
-            }
-        }
-        Err(e) => tracing::warn!(error = %e, "could not load control catalog for triage"),
-    }
-    specs
-}
-
-/// Absence-based control pass (the grounded half of the hybrid coverage): for each
-/// control with a [`surface`] definition, deterministically retrieve the code
-/// surfaces it governs (login routes, logging setup, update/download code) and have
-/// the grounded judge decide whether the control holds there. Returns net-new
-/// findings, each already tagged with its control and grounded to a real snippet.
-///
-/// The orchestrator runs this when `breakpilot.grounded_control_checks` is set
-/// (on by default). Validated live; it covers the 8 absence-based CRA controls
-/// (the judge decides presence/absence, grounded to a real snippet).
-pub async fn grounded_surface_findings(
-    config: &AgentConfig,
-    llm: Arc<LlmClient>,
-    repo_path: &Path,
-    repo_id: &str,
-) -> Vec<Finding> {
-    let Some(base_url) = config.breakpilot.base_url.clone() else {
-        return Vec::new();
-    };
-    let provider = OscalControlsProvider::new(
-        reqwest::Client::new(),
-        base_url,
-        config.breakpilot.token.clone(),
-        &config.breakpilot.snapshot_dir,
-    );
-    let specs = build_specs(&provider).await;
-    if specs.is_empty() {
-        return Vec::new();
-    }
-    let checker = GroundedControlChecker::new(LlmControlJudge::new(llm));
-
-    let mut out = Vec::new();
-    for surf in surface::SURFACES {
-        let Some(spec) = specs.get(surf.control_id) else {
-            continue; // catalog doesn't carry this control
-        };
-        let regions = surface::retrieve(repo_path, surf.terms);
-        if regions.is_empty() {
-            continue;
-        }
-        out.extend(checker.check(spec, &regions, repo_id).await);
-    }
-    out
-}
-
-/// Read a window of lines around `line` (1-based) from `repo_path/file`.
-fn fetch_region(repo_path: &Path, file: &str, line: u32) -> Option<CandidateRegion> {
-    let content = std::fs::read_to_string(repo_path.join(file)).ok()?;
-    let lines: Vec<&str> = content.lines().collect();
-    if lines.is_empty() {
-        return None;
-    }
-    let center = (line.saturating_sub(1) as usize).min(lines.len() - 1);
-    let start = center.saturating_sub(REGION_WINDOW);
-    let end = (center + REGION_WINDOW + 1).min(lines.len());
-    Some(CandidateRegion {
-        file: file.to_string(),
-        start_line: (start as u32) + 1,
-        content: lines[start..end].join("\n"),
-    })
-}
-
-/// Master-controls **semantic** pass: for each finding's code region, retrieve the
-/// top-K nearest master controls by embedding, have the grounded judge confirm,
-/// and stamp the confirmed control ids onto the finding — the scale path for the
-/// ~13.6k master-control corpus (which has no CWE to LUT on). Returns the number
-/// of findings that gained a master-control ref.
-///
-/// The orchestrator runs this when `breakpilot.semantic_mapping` is set (on by
-/// default). The control embedding index is built once and cached to
-/// `snapshot_dir` keyed by corpus hash ([`ControlIndex::load_or_build`]), so only
-/// the first scan after a catalog change pays the embedding cost.
-pub async fn semantic_stamp_findings(
-    config: &AgentConfig,
-    llm: Arc<LlmClient>,
-    repo_path: &Path,
-    findings: &mut [Finding],
-) -> usize {
-    let Some(base_url) = config.breakpilot.base_url.clone() else {
-        return 0;
-    };
-    let provider = OscalControlsProvider::new(
-        reqwest::Client::new(),
-        base_url,
-        config.breakpilot.token.clone(),
-        &config.breakpilot.snapshot_dir,
-    );
-    let doc = match provider.load_master_controls().await {
-        Ok(d) => d,
-        Err(e) => {
-            tracing::warn!(error = %e, "master-controls catalog unavailable; skipping semantic pass");
-            return 0;
-        }
-    };
-    let specs: Vec<ControlCheckSpec> = doc
-        .to_controls()
-        .into_iter()
-        .map(|c| ControlCheckSpec {
-            control_id: c.id,
-            title: c.title,
-            requirement: c.text,
-            default_cwe: None,
-            severity: Severity::Medium,
-        })
-        .collect();
-    let cache_path =
-        Path::new(&config.breakpilot.snapshot_dir).join("control-index-master-controls.json");
-    let index = match ControlIndex::load_or_build(&llm, specs, &cache_path).await {
-        Ok(i) if !i.is_empty() => i,
-        Ok(_) => return 0,
-        Err(e) => {
-            tracing::warn!(error = %e, "failed to embed master-controls corpus");
-            return 0;
-        }
-    };
-    let checker = SemanticControlChecker::new(LlmControlJudge::new(llm.clone()));
-
-    let mut tagged = 0;
-    for finding in findings.iter_mut() {
-        if finding.status == FindingStatus::FalsePositive {
-            continue;
-        }
-        let (Some(file), Some(line)) = (finding.file_path.clone(), finding.line_number) else {
-            continue;
-        };
-        let Some(region) = fetch_region(repo_path, &file, line) else {
-            continue;
-        };
-        // Retrieve on the finding's intent + the code, not the region alone: two
-        // findings in one file share overlapping windows and otherwise embed alike,
-        // collapsing onto the same controls. The finding's title/description carry
-        // the discriminating signal (e.g. "brute-force protection" vs "weak hash").
-        // The raw `region` still goes to the judge for snippet grounding.
-        let query = format!(
-            "{}\n{}\n\n{}",
-            finding.title, finding.description, region.content
-        );
-        let query_emb = match llm.embed(vec![query]).await {
-            Ok(mut embs) => match embs.pop() {
-                Some(v) => v,
-                None => continue,
-            },
-            Err(e) => {
-                tracing::warn!(error = %e, "query embed failed; skipping finding");
-                continue;
-            }
-        };
-        let confirmed = checker
-            .check(
-                &index,
-                &region,
-                &query_emb,
-                SEMANTIC_TOP_K,
-                &finding.repo_id,
-            )
-            .await;
-        let before = finding.control_refs.len();
-        for f in confirmed {
-            for cref in f.control_refs {
-                if !finding.control_refs.contains(&cref) {
-                    finding.control_refs.push(cref);
-                }
-            }
-        }
-        if finding.control_refs.len() > before {
-            tagged += 1;
-        }
-    }
-    tagged
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn fetch_region_windows_around_the_line() {
-        let dir = std::env::temp_dir().join(format!("triage-region-{}", uuid::Uuid::new_v4()));
-        std::fs::create_dir_all(&dir).unwrap();
-        let file = "a.py";
-        std::fs::write(dir.join(file), "l1\nl2\nl3\nSECRET=1\nl5\nl6\n").unwrap();
-        let r = fetch_region(&dir, file, 4).unwrap();
-        assert!(r.content.contains("SECRET=1"));
-        assert_eq!(r.start_line, 1); // window clamps to file start
-        let _ = std::fs::remove_dir_all(&dir);
-    }
-
-    #[test]
-    fn fetch_region_missing_file_is_none() {
-        assert!(fetch_region(Path::new("/nonexistent"), "nope.py", 1).is_none());
-    }
-}
@@ -1,121 +0,0 @@
-//! Semantic control mapping: retrieve the top-K controls nearest a code region,
-//! then confirm each with the grounded judge.
-//!
-//! The `region → controls` direction (vs. the CWE-LUT's `finding → control`) is
-//! what scales to the full master-control corpus: the LLM only ever judges a
-//! handful of retrieved candidates, and every surviving verdict is still anchored
-//! to real code by the grounding gate.
-
-use compliance_core::control_check::{ground, CandidateRegion};
-use compliance_core::models::Finding;
-
-use super::index::ControlIndex;
-use super::judge::ControlJudge;
-
-/// Retrieve → judge → ground, generic over the judge so tests use a stub.
-pub struct SemanticControlChecker<J> {
-    judge: J,
-}
-
-impl<J: ControlJudge> SemanticControlChecker<J> {
-    pub fn new(judge: J) -> Self {
-        Self { judge }
-    }
-
-    /// Map a code region to the controls it violates. `query_embedding` is the
-    /// caller-supplied retrieval embedding — typically the finding's intent
-    /// (title/description) plus the region, so retrieval keys on what the finding
-    /// is *about*, not just the ambient code. The top-`k` nearest controls in
-    /// `index` are then judged against the raw `region` and grounded.
-    pub async fn check(
-        &self,
-        index: &ControlIndex,
-        region: &CandidateRegion,
-        query_embedding: &[f64],
-        k: usize,
-        repo_id: &str,
-    ) -> Vec<Finding> {
-        let candidates = index.nearest(query_embedding, k);
-        let mut findings = Vec::new();
-        for spec in &candidates {
-            let verdict = self.judge.judge(spec, region).await;
-            if let Some(finding) = ground(spec, region, &verdict, repo_id) {
-                findings.push(finding);
-            }
-        }
-        findings
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use compliance_core::control_check::{ControlCheckSpec, LlmVerdict};
-    use compliance_core::models::finding::Severity;
-
-    struct StubJudge {
-        verdict: LlmVerdict,
-    }
-    impl ControlJudge for StubJudge {
-        async fn judge(&self, _s: &ControlCheckSpec, _r: &CandidateRegion) -> LlmVerdict {
-            self.verdict.clone()
-        }
-    }
-
-    fn spec(id: &str) -> ControlCheckSpec {
-        ControlCheckSpec {
-            control_id: id.into(),
-            title: id.into(),
-            requirement: id.into(),
-            default_cwe: None,
-            severity: Severity::Medium,
-        }
-    }
-
-    #[tokio::test]
-    async fn retrieves_then_grounds_the_nearest_control() {
-        let index = ControlIndex::from_embeddings(vec![
-            (spec("mc-near"), vec![1.0, 0.0]),
-            (spec("mc-far"), vec![0.0, 1.0]),
-        ]);
-        let checker = SemanticControlChecker::new(StubJudge {
-            verdict: LlmVerdict {
-                violates: true,
-                snippet: "PASSWORD = \"admin\"".into(),
-                cwe: None,
-                confidence: 0.9,
-            },
-        });
-        let region = CandidateRegion {
-            file: "src/auth.py".into(),
-            start_line: 1,
-            content: "PASSWORD = \"admin\"\n".into(),
-        };
-        // Query embedding nearest to mc-near; k=1 → only mc-near is judged.
-        let findings = checker
-            .check(&index, &region, &[0.95, 0.05], 1, "repo")
-            .await;
-        assert_eq!(findings.len(), 1);
-        assert_eq!(findings[0].control_refs, vec!["mc-near".to_string()]);
-    }
-
-    #[tokio::test]
-    async fn ungrounded_verdict_is_dropped() {
-        let index = ControlIndex::from_embeddings(vec![(spec("mc-near"), vec![1.0, 0.0])]);
-        let checker = SemanticControlChecker::new(StubJudge {
-            verdict: LlmVerdict {
-                violates: true,
-                snippet: "not in the region".into(),
-                cwe: None,
-                confidence: 0.9,
-            },
-        });
-        let region = CandidateRegion {
-            file: "f".into(),
-            start_line: 1,
-            content: "real code\n".into(),
-        };
-        let findings = checker.check(&index, &region, &[1.0, 0.0], 1, "repo").await;
-        assert!(findings.is_empty());
-    }
-}
@@ -1,258 +0,0 @@
-//! Surface retrieval for absence-based controls.
-//!
-//! Some CRA controls are violated by an *absence* — no rate limiting on login, no
-//! security logging, no signature check on an update — so there's no offending
-//! pattern for semgrep to match. Instead we deterministically locate the code
-//! *surface* the control governs (a login route, a logging setup, update/download
-//! code) by identifier/route terms, then hand each surface region to the grounded
-//! judge, which decides whether the control is satisfied there. The resulting
-//! finding grounds to the surface snippet, so nothing fabricated survives.
-//!
-//! Retrieval is intentionally cheap and bounded: keyword match + a fixed window,
-//! capped per control to keep the downstream LLM cost predictable.
-
-use std::path::Path;
-
-use compliance_core::control_check::CandidateRegion;
-
-/// An absence-based control and the case-insensitive terms that mark the code
-/// surface it governs.
-pub struct Surface {
-    pub control_id: &'static str,
-    pub terms: &'static [&'static str],
-}
-
-/// The absence-based CRA controls we retrieve surfaces for — the grounded half of
-/// the hybrid coverage (the pattern-expressible half is custom semgrep rules).
-pub const SURFACES: &[Surface] = &[
-    Surface {
-        control_id: "cra-ai-6", // Integritaetspruefung
-        terms: &[
-            "checksum",
-            "sha256",
-            "signature",
-            "hmac",
-            "integrity",
-            "verify",
-        ],
-    },
-    Surface {
-        control_id: "cra-ai-11", // Brute-Force-Schutz
-        terms: &[
-            "login",
-            "signin",
-            "authenticate",
-            "/auth",
-            "password",
-            "ratelimit",
-        ],
-    },
-    Surface {
-        control_id: "cra-ai-12", // Rollenbasierte Autorisierung (RBAC)
-        terms: &[
-            "authorize",
-            "permission",
-            "role",
-            "rbac",
-            "require_role",
-            "has_role",
-        ],
-    },
-    Surface {
-        control_id: "cra-ai-24", // Security-Logging
-        terms: &["login", "authorize", "permission", "role", "admin", "audit"],
-    },
-    Surface {
-        control_id: "cra-ai-27", // Log-Integritaet und -Aufbewahrung
-        terms: &["logging", "logger", "getlogger", "audit_log"],
-    },
-    Surface {
-        control_id: "cra-ai-28", // Sichere Update-Mechanismen
-        terms: &["update", "upgrade", "download", "firmware"],
-    },
-    Surface {
-        control_id: "cra-ai-29", // Update-Authentizitaet
-        terms: &["update", "signature", "verify", "pubkey", "certificate"],
-    },
-    Surface {
-        control_id: "cra-ai-30", // Update-Integritaet
-        terms: &["update", "checksum", "digest", "integrity", "verify"],
-    },
-];
-
-/// Source file extensions worth reading (skip binaries/assets/lockfiles).
-const CODE_EXTS: &[&str] = &[
-    "py", "js", "ts", "tsx", "jsx", "go", "java", "rb", "php", "rs", "cs", "kt",
-];
-/// Directories never worth walking.
-const SKIP_DIRS: &[&str] = &[
-    ".git",
-    "node_modules",
-    "target",
-    "vendor",
-    ".venv",
-    "__pycache__",
-    "dist",
-    "build",
-];
-/// Lines of context on each side of a hit.
-const WINDOW: usize = 6;
-/// Cap on regions per control, to bound downstream LLM calls.
-const MAX_REGIONS_PER_CONTROL: usize = 8;
-/// Skip files larger than this (generated/minified).
-const MAX_FILE_BYTES: u64 = 512 * 1024;
-
-/// Deterministically retrieve up to [`MAX_REGIONS_PER_CONTROL`] code regions in
-/// `repo_path` whose lines mention any of `terms`. Hits close together within a
-/// file are merged into one region; results are capped to bound LLM cost.
-pub fn retrieve(repo_path: &Path, terms: &[&str]) -> Vec<CandidateRegion> {
-    let lowered: Vec<String> = terms.iter().map(|t| t.to_lowercase()).collect();
-    let mut regions = Vec::new();
-    for entry in walk(repo_path) {
-        if regions.len() >= MAX_REGIONS_PER_CONTROL {
-            break;
-        }
-        let path = entry.path();
-        if !has_code_ext(path) {
-            continue;
-        }
-        let Ok(meta) = entry.metadata() else { continue };
-        if !meta.is_file() || meta.len() > MAX_FILE_BYTES {
-            continue;
-        }
-        let Ok(content) = std::fs::read_to_string(path) else {
-            continue;
-        };
-        let rel = path
-            .strip_prefix(repo_path)
-            .unwrap_or(path)
-            .to_string_lossy()
-            .to_string();
-        let lines: Vec<&str> = content.lines().collect();
-        let hits: Vec<usize> = lines
-            .iter()
-            .enumerate()
-            .filter(|(_, line)| {
-                let ll = line.to_lowercase();
-                lowered.iter().any(|t| ll.contains(t.as_str()))
-            })
-            .map(|(i, _)| i)
-            .collect();
-        for center in merge_centers(&hits) {
-            if regions.len() >= MAX_REGIONS_PER_CONTROL {
-                break;
-            }
-            let start = center.saturating_sub(WINDOW);
-            let end = (center + WINDOW + 1).min(lines.len());
-            regions.push(CandidateRegion {
-                file: rel.clone(),
-                start_line: (start as u32) + 1,
-                content: lines[start..end].join("\n"),
-            });
-        }
-    }
-    regions
-}
-
-/// Collapse ascending hit indices that fall within one window into a single
-/// representative center, so overlapping regions aren't judged repeatedly.
-fn merge_centers(hits: &[usize]) -> Vec<usize> {
-    let mut out: Vec<usize> = Vec::new();
-    for &h in hits {
-        match out.last() {
-            Some(&last) if h.saturating_sub(last) <= WINDOW => {}
-            _ => out.push(h),
-        }
-    }
-    out
-}
-
-fn has_code_ext(path: &Path) -> bool {
-    path.extension()
-        .and_then(|e| e.to_str())
-        .is_some_and(|e| CODE_EXTS.contains(&e))
-}
-
-fn walk(root: &Path) -> Vec<walkdir::DirEntry> {
-    walkdir::WalkDir::new(root)
-        .into_iter()
-        .filter_entry(|e| {
-            let name = e.file_name().to_string_lossy();
-            !SKIP_DIRS.contains(&name.as_ref())
-        })
-        .filter_map(|e| e.ok())
-        .collect()
-}
-
-#[cfg(test)]
-#[allow(clippy::unwrap_used)]
-mod tests {
-    use super::*;
-
-    fn write(dir: &Path, rel: &str, body: &str) {
-        let p = dir.join(rel);
-        if let Some(parent) = p.parent() {
-            std::fs::create_dir_all(parent).unwrap();
-        }
-        std::fs::write(p, body).unwrap();
-    }
-
-    fn terms_for(control_id: &str) -> &'static [&'static str] {
-        SURFACES
-            .iter()
-            .find(|s| s.control_id == control_id)
-            .unwrap()
-            .terms
-    }
-
-    #[test]
-    fn surfaces_cover_the_absence_based_controls() {
-        assert_eq!(SURFACES.len(), 8);
-        for id in [
-            "cra-ai-6",
-            "cra-ai-11",
-            "cra-ai-12",
-            "cra-ai-24",
-            "cra-ai-27",
-            "cra-ai-28",
-            "cra-ai-29",
-            "cra-ai-30",
-        ] {
-            assert!(SURFACES.iter().any(|s| s.control_id == id), "{id} missing");
-        }
-    }
-
-    #[test]
-    fn retrieves_matching_region_with_context() {
-        let dir = std::env::temp_dir().join(format!("surface-{}", uuid::Uuid::new_v4()));
-        std::fs::create_dir_all(&dir).unwrap();
-        write(
-            &dir,
-            "app/auth.py",
-            "import x\n\n\n\n\n\n\ndef login(u, p):\n    return check(u, p)\n",
-        );
-        let regions = retrieve(&dir, terms_for("cra-ai-11"));
-        assert_eq!(regions.len(), 1);
-        assert!(regions[0].content.contains("def login"));
-        assert_eq!(regions[0].file, "app/auth.py");
-        let _ = std::fs::remove_dir_all(&dir);
-    }
-
-    #[test]
-    fn skips_non_code_and_vendored() {
-        let dir = std::env::temp_dir().join(format!("surface-{}", uuid::Uuid::new_v4()));
-        std::fs::create_dir_all(&dir).unwrap();
-        write(&dir, "README.md", "login and password and audit\n"); // not code ext
-        write(&dir, "node_modules/pkg/index.js", "function login() {}\n"); // vendored
-        assert!(retrieve(&dir, terms_for("cra-ai-11")).is_empty());
-        let _ = std::fs::remove_dir_all(&dir);
-    }
-
-    #[test]
-    fn merges_adjacent_hits_into_one_region() {
-        // Two hits one line apart collapse to a single center/region.
-        assert_eq!(merge_centers(&[10, 11, 30]), vec![10, 30]);
-        assert_eq!(merge_centers(&[]), Vec::<usize>::new());
-        assert_eq!(merge_centers(&[5]), vec![5]);
-    }
-}
@@ -1,234 +0,0 @@
-//! Triage step: confirm/refute a deterministic tool finding against the controls
-//! it maps to (via the `control-map` LUT), grounding the judgment.
-//!
-//! This is where the LLM finally enters — as a **false-positive filter over tool
-//! output**, never as the detector (the ZeroFalse / IRIS pattern). A tool
-//! (semgrep, gitleaks, syft/osv) detects deterministically; `controls_for(tool,
-//! cwe)` attaches the finding to the control(s) it's evidence for; the grounded
-//! judge then confirms or refutes each, and only judgments anchored to real code
-//! survive.
-
-use std::collections::HashMap;
-
-use compliance_core::control_check::{ground, CandidateRegion, ControlCheckSpec};
-use compliance_core::models::Finding;
-use control_map::ControlMap;
-
-use super::judge::ControlJudge;
-
-/// What triage decided for one tool finding.
-#[derive(Debug, Clone, PartialEq, Eq)]
-pub enum TriageOutcome {
-    /// The finding maps to no control in the LUT — keep it, untagged.
-    Unmapped,
-    /// Maps to controls and the grounded judge confirmed at least one — keep the
-    /// finding and tag it with these control ids.
-    Confirmed(Vec<String>),
-    /// Maps to controls but the judge grounded none — treat as a false positive.
-    FalsePositive,
-}
-
-/// Triages tool findings against the control map, confirming with a grounded judge.
-pub struct ControlTriage<J> {
-    judge: J,
-    map: ControlMap,
-    /// Control requirement specs (by control id), built from the ingested catalog.
-    specs: HashMap<String, ControlCheckSpec>,
-}
-
-impl<J: ControlJudge> ControlTriage<J> {
-    pub fn new(judge: J, map: ControlMap, specs: HashMap<String, ControlCheckSpec>) -> Self {
-        Self { judge, map, specs }
-    }
-
-    /// Triage one tool finding. `region` is the code around the finding, used as
-    /// the grounding evidence for the judge.
-    pub async fn triage(&self, finding: &Finding, region: &CandidateRegion) -> TriageOutcome {
-        // Match by CWE (off-the-shelf findings) and/or rule id (our custom
-        // detectors, which carry no LUT-bound CWE). A finding with neither is
-        // simply unmapped.
-        let mapped = self.map.controls_for_finding(
-            &finding.scanner,
-            finding.cwe.as_deref(),
-            finding.rule_id.as_deref(),
-        );
-        if mapped.is_empty() {
-            return TriageOutcome::Unmapped;
-        }
-
-        let mut confirmed = Vec::new();
-        for entry in mapped {
-            let Some(spec) = self.specs.get(&entry.control) else {
-                continue;
-            };
-            let verdict = self.judge.judge(spec, region).await;
-            // The verdict only counts if it grounds to real code in the region.
-            if ground(spec, region, &verdict, &finding.repo_id).is_some() {
-                confirmed.push(entry.control.clone());
-            }
-        }
-
-        if confirmed.is_empty() {
-            TriageOutcome::FalsePositive
-        } else {
-            TriageOutcome::Confirmed(confirmed)
-        }
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use compliance_core::control_check::LlmVerdict;
-    use compliance_core::models::finding::Severity;
-    use compliance_core::models::scan::ScanType;
-
-    struct StubJudge {
-        verdict: LlmVerdict,
-    }
-    impl ControlJudge for StubJudge {
-        async fn judge(&self, _s: &ControlCheckSpec, _r: &CandidateRegion) -> LlmVerdict {
-            self.verdict.clone()
-        }
-    }
-
-    fn specs() -> HashMap<String, ControlCheckSpec> {
-        let mut m = HashMap::new();
-        m.insert(
-            "cra-ai-8".to_string(),
-            ControlCheckSpec {
-                control_id: "cra-ai-8".into(),
-                title: "No default passwords".into(),
-                requirement: "No default credentials".into(),
-                default_cwe: Some("CWE-798".into()),
-                severity: Severity::High,
-            },
-        );
-        m
-    }
-
-    fn semgrep_finding(cwe: &str) -> Finding {
-        let mut f = Finding::new(
-            "repo".into(),
-            "fp1".into(),
-            "semgrep".into(),
-            ScanType::Sast,
-            "hardcoded credential".into(),
-            "desc".into(),
-            Severity::High,
-        );
-        f.cwe = Some(cwe.into());
-        f
-    }
-
-    fn region() -> CandidateRegion {
-        CandidateRegion {
-            file: "src/auth.py".into(),
-            start_line: 1,
-            content: "PASSWORD = \"admin123\"\n".into(),
-        }
-    }
-
-    #[tokio::test]
-    async fn confirmed_finding_is_tagged_with_control() {
-        let triage = ControlTriage::new(
-            StubJudge {
-                verdict: LlmVerdict {
-                    violates: true,
-                    snippet: "PASSWORD = \"admin123\"".into(),
-                    cwe: None,
-                    confidence: 0.9,
-                },
-            },
-            ControlMap::cra().unwrap(),
-            specs(),
-        );
-        let out = triage.triage(&semgrep_finding("CWE-798"), &region()).await;
-        assert_eq!(out, TriageOutcome::Confirmed(vec!["cra-ai-8".to_string()]));
-    }
-
-    #[tokio::test]
-    async fn refuted_mapped_finding_is_false_positive() {
-        // Maps to cra-ai-8, but the judge doesn't confirm (no violation) → FP.
-        let triage = ControlTriage::new(
-            StubJudge {
-                verdict: LlmVerdict {
-                    violates: false,
-                    snippet: String::new(),
-                    cwe: None,
-                    confidence: 0.1,
-                },
-            },
-            ControlMap::cra().unwrap(),
-            specs(),
-        );
-        let out = triage.triage(&semgrep_finding("CWE-798"), &region()).await;
-        assert_eq!(out, TriageOutcome::FalsePositive);
-    }
-
-    #[tokio::test]
-    async fn custom_rule_finding_without_cwe_is_confirmed() {
-        // A custom detector finding carries a rule id but no LUT-bound CWE; it must
-        // still map (by rule id) and confirm.
-        let mut specs = specs();
-        specs.insert(
-            "cra-ai-1".to_string(),
-            ControlCheckSpec {
-                control_id: "cra-ai-1".into(),
-                title: "Secure-by-Default".into(),
-                requirement: "Ship secure defaults".into(),
-                default_cwe: None,
-                severity: Severity::Medium,
-            },
-        );
-        let triage = ControlTriage::new(
-            StubJudge {
-                verdict: LlmVerdict {
-                    violates: true,
-                    snippet: "app.run(debug=True)".into(),
-                    cwe: None,
-                    confidence: 0.9,
-                },
-            },
-            ControlMap::cra().unwrap(),
-            specs,
-        );
-        let mut f = Finding::new(
-            "repo".into(),
-            "fp".into(),
-            "semgrep".into(),
-            ScanType::Sast,
-            "flask debug".into(),
-            "desc".into(),
-            Severity::Medium,
-        );
-        f.rule_id = Some("tmp.x.cra-ai-1-flask-debug-enabled".into()); // no cwe
-        let region = CandidateRegion {
-            file: "app.py".into(),
-            start_line: 1,
-            content: "app.run(debug=True)\n".into(),
-        };
-        let out = triage.triage(&f, &region).await;
-        assert_eq!(out, TriageOutcome::Confirmed(vec!["cra-ai-1".to_string()]));
-    }
-
-    #[tokio::test]
-    async fn unmapped_cwe_is_left_untagged() {
-        let triage = ControlTriage::new(
-            StubJudge {
-                verdict: LlmVerdict {
-                    violates: true,
-                    snippet: "PASSWORD = \"admin123\"".into(),
-                    cwe: None,
-                    confidence: 0.9,
-                },
-            },
-            ControlMap::cra().unwrap(),
-            specs(),
-        );
-        let out = triage
-            .triage(&semgrep_finding("CWE-99999"), &region())
-            .await;
-        assert_eq!(out, TriageOutcome::Unmapped);
-    }
-}
@@ -1,233 +1,11 @@
-use std::sync::Arc;
-
-use dashmap::DashMap;
 use mongodb::bson::doc;
 use mongodb::options::IndexOptions;
 use mongodb::{Client, Collection, IndexModel};
-use sha2::{Digest, Sha256};

 use compliance_core::models::*;
-use compliance_core::TenantContext;

 use crate::error::AgentError;

-/// Mongo enforces a 63-byte cap on database names (older clusters: 64
-/// on Linux, 63 on Windows; we target the conservative limit).
-const MAX_DB_NAME_LEN: usize = 63;
-
-/// Hex length of the SHA-256 truncation used for the hash fallback
-/// tenant DB name (16 bytes → 32 hex chars). 16 bytes gives ~2^64
-/// birthday-collision resistance — at our 10s-100s tenant scale this
-/// is effectively impossible to hit.
-const HASH_HEX_LEN: usize = 32;
-
-/// Largest `db_prefix` that still guarantees the hash-fallback name
-/// fits in the 63-byte cap: `prefix + "_" + 32 hex chars`.
-const MAX_PREFIX_LEN: usize = MAX_DB_NAME_LEN - 1 - HASH_HEX_LEN;
-
-/// Per-tenant Mongo connection broker (M7.2 isolation model).
-///
-/// Holds one [`Client`] and hands out [`Database`] handles physically
-/// scoped to `<db_prefix>_<tenant_id>`. The driver is the isolation
-/// boundary — a handle for tenant A cannot see tenant B's documents
-/// because it is connected to a different database, not because of an
-/// application-level filter.
-///
-/// Index creation runs idempotently the first time each tenant is seen
-/// in the process's lifetime. Mongo's `createIndex` is itself idempotent
-/// by index name; the in-memory `ensured` set just skips the round-trip.
-#[derive(Clone, Debug)]
-pub struct DatabasePool {
-    client: Client,
-    db_prefix: String,
-    ensured: Arc<DashMap<String, ()>>,
-}
-
-impl DatabasePool {
-    /// Connect to the cluster and prepare to hand out tenant databases
-    /// named `<db_prefix>_<tenant_id>`.
-    ///
-    /// Validates `db_prefix.len() <= MAX_PREFIX_LEN` so the
-    /// hash-fallback path is provably within Mongo's 63-byte db-name
-    /// cap. Refuses to construct a pool that could ever produce an
-    /// over-long name.
-    pub async fn connect(uri: &str, db_prefix: &str) -> Result<Self, AgentError> {
-        if db_prefix.len() > MAX_PREFIX_LEN {
-            return Err(AgentError::Other(format!(
-                "db_prefix '{db_prefix}' is {} chars; max is {MAX_PREFIX_LEN} so the \
-                 hash-fallback tenant DB name fits Mongo's {MAX_DB_NAME_LEN}-byte cap",
-                db_prefix.len()
-            )));
-        }
-        let client = Client::with_uri_str(uri).await?;
-        client
-            .database("admin")
-            .run_command(doc! { "ping": 1 })
-            .await?;
-        tracing::info!(
-            "MongoDB cluster reachable; per-tenant pool ready (db prefix '{db_prefix}')"
-        );
-        Ok(Self {
-            client,
-            db_prefix: db_prefix.to_string(),
-            ensured: Arc::new(DashMap::new()),
-        })
-    }
-
-    /// Return a [`Database`] scoped to this tenant. Ensures indexes on
-    /// first call per tenant (per process). Cheap on the hot path —
-    /// subsequent calls skip the round-trip.
-    pub async fn for_tenant(&self, ctx: &TenantContext) -> Result<Database, AgentError> {
-        self.for_tenant_id(&ctx.tenant_id).await
-    }
-
-    /// Like [`Self::for_tenant`] but accepts a bare tenant_id.
-    /// For background paths (scheduler, webhooks, pipeline orchestrators)
-    /// that don't have a full [`TenantContext`] but know which tenant
-    /// they're operating on (typically resolved from a URL path, a job
-    /// argument, or the registry).
-    pub async fn for_tenant_id(&self, tenant_id: &str) -> Result<Database, AgentError> {
-        let db_name = self.tenant_db_name(tenant_id);
-        let db = Database::from_database(self.client.database(&db_name));
-        // `DashMap::insert` returns the previous value; `None` means we
-        // were the first writer for this tenant_id and own the
-        // index-ensure work.
-        if self.ensured.insert(tenant_id.to_string(), ()).is_none() {
-            if let Err(e) = db.ensure_indexes().await {
-                // Roll the marker back so the next request retries.
-                self.ensured.remove(tenant_id);
-                return Err(e);
-            }
-            tracing::debug!(
-                tenant_id = %tenant_id,
-                db_name = %db_name,
-                "Indexes ensured for tenant database"
-            );
-        }
-        Ok(db)
-    }
-
-    /// Compute the Mongo database name for a tenant. Public for tests
-    /// and tenant offboarding (`pool.client().database(name).drop()`).
-    ///
-    /// Format: `<prefix>_<sanitized_tenant_id>` if it fits the 63-byte
-    /// cap, else `<prefix>_<sha256-16-byte-hex-of-tenant_id>`. The
-    /// `db_prefix` length invariant established at [`Self::connect`]
-    /// guarantees the hash-fallback name always fits — no runtime
-    /// assertion needed.
-    ///
-    /// Collision resistance: the hash fallback is a 16-byte SHA-256
-    /// truncation, which gives ~2^64 birthday-collision resistance. At
-    /// our 10s–100s tenant scale the probability of two tenant_ids
-    /// colliding is effectively zero. (8-byte truncation would have
-    /// been ~2^32 — too close for comfort on a regulated product.)
-    pub fn tenant_db_name(&self, tenant_id: &str) -> String {
-        let sanitized = sanitize_tenant_id(tenant_id);
-        let natural = format!("{}_{}", self.db_prefix, sanitized);
-        if natural.len() <= MAX_DB_NAME_LEN {
-            natural
-        } else {
-            let mut hasher = Sha256::new();
-            hasher.update(tenant_id.as_bytes());
-            let digest = hasher.finalize();
-            let suffix = hex::encode(&digest[..HASH_HEX_LEN / 2]);
-            format!("{}_{}", self.db_prefix, suffix)
-        }
-    }
-
-    /// Raw client handle. Reserved for cross-tenant admin flows that
-    /// must opt in explicitly (tenant listing, drop-on-offboard).
-    pub fn client(&self) -> &Client {
-        &self.client
-    }
-
-    /// Cross-tenant admin database used by features that intentionally
-    /// span tenants (today: MCP bearer tokens — each token row carries
-    /// a `tenant_id` and the MCP server reads them to route requests).
-    ///
-    /// The name `<db_prefix>__admin` (double underscore) is reserved —
-    /// the sanitizer never produces it for a normal tenant DB because
-    /// the natural format is `<db_prefix>_<sanitized_tenant_id>` (one
-    /// underscore) and tenant_ids would have to start with `_admin` to
-    /// collide. New tenant provisioning should reject such ids.
-    pub fn admin_db(&self) -> mongodb::Database {
-        self.client.database(&self.admin_db_name())
-    }
-
-    /// Name of the admin database — public so tests / operators can
-    /// drop it via the raw client.
-    pub fn admin_db_name(&self) -> String {
-        format!("{}__admin", self.db_prefix)
-    }
-
-    /// List every Mongo database currently belonging to this pool,
-    /// identified by the `<db_prefix>_` prefix. The result is the raw
-    /// database names — opening one for offboarding/cleanup goes
-    /// through [`Self::client`].
-    ///
-    /// Note: hashed-fallback names (very long tenant_ids) lose the
-    /// original tenant_id at the cluster level — we know a database
-    /// exists for *some* tenant but not which one. In practice
-    /// tenant_ids are UUIDs (36 chars) and never hit the fallback,
-    /// so this is a theoretical concern, not an operational one.
-    pub async fn list_tenant_db_names(&self) -> Result<Vec<String>, AgentError> {
-        let prefix = format!("{}_", self.db_prefix);
-        let names = self.client.list_database_names().await?;
-        Ok(names
-            .into_iter()
-            .filter(|n| n.starts_with(&prefix))
-            .collect())
-    }
-
-    /// Tenant ids for every provisioned tenant database, derived by stripping
-    /// the `<prefix>_` from the database names. Skips the admin database
-    /// (`<prefix>__admin`). Hash-fallback names (very long tenant_ids) are lost
-    /// at the cluster level and cannot be recovered here — in practice tenant
-    /// ids are UUIDs and never hit that path. Used by the migration CLI's
-    /// `--all` mode.
-    pub async fn list_tenant_ids(&self) -> Result<Vec<String>, AgentError> {
-        let prefix = format!("{}_", self.db_prefix);
-        Ok(self
-            .list_tenant_db_names()
-            .await?
-            .into_iter()
-            .filter_map(|n| n.strip_prefix(&prefix).map(str::to_string))
-            .filter(|id| !id.starts_with('_'))
-            .collect())
-    }
-
-    /// Drop the database for a specific tenant. Used by GDPR delete
-    /// and tenant offboarding. Idempotent — dropping a non-existent
-    /// database is a no-op at the driver level.
-    ///
-    /// Also evicts the tenant from the in-memory `ensured` set so a
-    /// later re-provision triggers fresh `ensure_indexes`.
-    pub async fn drop_tenant(&self, tenant_id: &str) -> Result<(), AgentError> {
-        let db_name = self.tenant_db_name(tenant_id);
-        self.client.database(&db_name).drop().await?;
-        self.ensured.remove(tenant_id);
-        tracing::info!(
-            tenant_id = %tenant_id,
-            db_name = %db_name,
-            "Dropped tenant database"
-        );
-        Ok(())
-    }
-}
-
-/// Mongo database names disallow `/`, `\`, `.`, `"`, `$`, ` `, and NUL.
-/// breakpilot-dev tenant_ids are UUIDs so this is belt-and-braces, but
-/// it lets the pool tolerate any future tenant_id shape without surprise.
-fn sanitize_tenant_id(tenant_id: &str) -> String {
-    tenant_id
-        .chars()
-        .map(|c| match c {
-            '/' | '\\' | '.' | '"' | '$' | ' ' | '\0' => '_',
-            c => c,
-        })
-        .collect()
-}
-
 #[derive(Clone, Debug)]
 pub struct Database {
    inner: mongodb::Database,
@@ -242,13 +20,17 @@ impl Database {
        Ok(Self { inner: db })
    }

-    /// Wrap an already-resolved Mongo database. Used by [`DatabasePool`]
-    /// to hand out tenant-scoped handles without a fresh client per tenant.
-    pub(crate) fn from_database(inner: mongodb::Database) -> Self {
-        Self { inner }
-    }
-
    pub async fn ensure_indexes(&self) -> Result<(), AgentError> {
+        // repositories: unique git_url
+        self.repositories()
+            .create_index(
+                IndexModel::builder()
+                    .keys(doc! { "git_url": 1 })
+                    .options(IndexOptions::builder().unique(true).build())
+                    .build(),
+            )
+            .await?;
+
        // findings: unique fingerprint
        self.findings()
            .create_index(
@@ -296,25 +78,6 @@ impl Database {
            )
            .await?;

-        // cve_notifications: unique cve_id + repo_id + package, status filter
-        self.cve_notifications()
-            .create_index(
-                IndexModel::builder()
-                    .keys(
-                        doc! { "cve_id": 1, "repo_id": 1, "package_name": 1, "package_version": 1 },
-                    )
-                    .options(IndexOptions::builder().unique(true).build())
-                    .build(),
-            )
-            .await?;
-        self.cve_notifications()
-            .create_index(
-                IndexModel::builder()
-                    .keys(doc! { "status": 1, "created_at": -1 })
-                    .build(),
-            )
-            .await?;
-
        // tracker_issues: unique finding_id
        self.tracker_issues()
            .create_index(
@@ -435,68 +198,14 @@ impl Database {
            )
            .await?;

-        // onboarded_targets: multikey on artifact source ref (webhook + dedupe
-        // lookup). Non-unique — "one git URL per tenant" is enforced in the
-        // create handler, since a unique multikey index on an array field has
-        // null-collision caveats.
-        self.onboarded_targets()
-            .create_index(
-                IndexModel::builder()
-                    .keys(doc! { "artifacts.source_ref": 1 })
-                    .build(),
-            )
-            .await?;
-
-        // onboarded_targets: multikey on artifact kind
-        self.onboarded_targets()
-            .create_index(
-                IndexModel::builder()
-                    .keys(doc! { "artifacts.kind": 1 })
-                    .build(),
-            )
-            .await?;
-
-        // onboarded_targets: target_type filter
-        self.onboarded_targets()
-            .create_index(
-                IndexModel::builder()
-                    .keys(doc! { "target_type": 1 })
-                    .build(),
-            )
-            .await?;
-
-        // werkbank_jobs: unique job id (idempotent enqueue by job id)
-        self.werkbank_jobs()
-            .create_index(
-                IndexModel::builder()
-                    .keys(doc! { "job.id": 1 })
-                    .options(IndexOptions::builder().unique(true).build())
-                    .build(),
-            )
-            .await?;
-
-        // werkbank_jobs: lease query — oldest queued job for an executor
-        self.werkbank_jobs()
-            .create_index(
-                IndexModel::builder()
-                    .keys(doc! { "status": 1, "job.executor": 1, "created_at": 1 })
-                    .build(),
-            )
-            .await?;
-
-        // werkbank_jobs: visibility-timeout sweep of expired leases
-        self.werkbank_jobs()
-            .create_index(
-                IndexModel::builder()
-                    .keys(doc! { "status": 1, "lease_expires_at": 1 })
-                    .build(),
-            )
-            .await?;
-
        tracing::info!("Database indexes ensured");
        Ok(())
    }

+    pub fn repositories(&self) -> Collection<TrackedRepository> {
+        self.inner.collection("repositories")
+    }
+
    pub fn findings(&self) -> Collection<Finding> {
        self.inner.collection("findings")
    }
@@ -513,12 +222,6 @@ impl Database {
        self.inner.collection("cve_alerts")
    }

-    pub fn cve_notifications(
-        &self,
-    ) -> Collection<compliance_core::models::notification::CveNotification> {
-        self.inner.collection("cve_notifications")
-    }
-
    pub fn tracker_issues(&self) -> Collection<TrackerIssue> {
        self.inner.collection("tracker_issues")
    }
@@ -545,20 +248,6 @@ impl Database {
        self.inner.collection("dast_targets")
    }

-    /// The unified onboarding targets that replace `repositories` and
-    /// `dast_targets`. Ids are preserved from the legacy collections during
-    /// migration so downstream `repo_id` / `target_id` references keep resolving.
-    pub fn onboarded_targets(&self) -> Collection<OnboardedTarget> {
-        self.inner.collection("onboarded_targets")
-    }
-
-    /// A typed handle to an arbitrary collection by name. For bookkeeping
-    /// collections without a dedicated model (e.g. `schema_migrations`,
-    /// `onboarding_migration_log`).
-    pub fn collection_named<T: Send + Sync>(&self, name: &str) -> Collection<T> {
-        self.inner.collection(name)
-    }
-
    pub fn dast_scan_runs(&self) -> Collection<DastScanRun> {
        self.inner.collection("dast_scan_runs")
    }
@@ -591,12 +280,6 @@ impl Database {
        self.inner.collection("pentest_messages")
    }

-    /// The Werkbank job queue (WB-02): declarative dynamic-execution jobs the
-    /// control plane enqueues and runners lease.
-    pub fn werkbank_jobs(&self) -> Collection<compliance_core::models::werkbank::JobRecord> {
-        self.inner.collection("werkbank_jobs")
-    }
-
    #[allow(dead_code)]
    pub fn raw_collection(&self, name: &str) -> Collection<mongodb::bson::Document> {
        self.inner.collection(name)
@@ -27,9 +27,6 @@ pub enum AgentError {
    #[error("Configuration error: {0}")]
    Config(String),

-    #[error("Dynamic-execution error: {0}")]
-    Exec(#[from] werkbank_exec::ExecError),
-
    #[error("{0}")]
    Other(String),
 }
@@ -1,178 +0,0 @@
-//! Content-addressed blob storage and archive extraction for ingest.
-//!
-//! Blobs are stored at `<base>/blobs/<sha[0:2]>/<sha>` and deduplicated by
-//! digest; per-run working directories live under `<base>/work/`.
-
-use std::fs::{self, File};
-use std::io::{self, Read};
-use std::path::{Path, PathBuf};
-
-use sha2::{Digest, Sha256};
-
-use crate::error::AgentError;
-
-/// Read buffer size for streaming hashes/copies (64 KiB).
-const BUF_LEN: usize = 64 * 1024;
-
-/// Stream-hash a file with SHA-256, returning the lowercase-hex digest and the
-/// byte length. Streams so large firmware images never load fully into memory.
-pub fn hash_file(path: &Path) -> Result<(String, u64), AgentError> {
-    let mut file = File::open(path)?;
-    let mut hasher = Sha256::new();
-    let mut buf = [0u8; BUF_LEN];
-    let mut total: u64 = 0;
-    loop {
-        let n = file.read(&mut buf)?;
-        if n == 0 {
-            break;
-        }
-        hasher.update(&buf[..n]);
-        total += n as u64;
-    }
-    Ok((hex::encode(hasher.finalize()), total))
-}
-
-/// Store raw bytes in the content-addressed blob store under `base`, returning
-/// the SHA-256 digest. Used to stash a small derived artifact (e.g. the extracted
-/// PLC program source) so a Werkbank runner can fetch it by hash. Idempotent.
-pub fn store_bytes(base: &Path, bytes: &[u8]) -> Result<String, AgentError> {
-    let sha = hex::encode(Sha256::digest(bytes));
-    let dir = base.join("blobs").join(&sha[0..2]);
-    fs::create_dir_all(&dir)?;
-    let dest = dir.join(&sha);
-    if !dest.exists() {
-        fs::write(&dest, bytes)?;
-    }
-    Ok(sha)
-}
-
-/// Read a blob's bytes by its SHA-256 digest. Rejects a non-hex/wrong-length hash
-/// so a request can't traverse outside the blob store.
-pub fn read_blob(base: &Path, sha: &str) -> Result<Vec<u8>, AgentError> {
-    if sha.len() != 64 || !sha.bytes().all(|b| b.is_ascii_hexdigit()) {
-        return Err(AgentError::Other(format!("invalid content hash '{sha}'")));
-    }
-    let path = base.join("blobs").join(&sha[0..2]).join(sha);
-    Ok(fs::read(path)?)
-}
-
-/// Copy `src` into the content-addressed blob store under `base`, returning the
-/// stored path. Idempotent: an already-present blob is not rewritten.
-pub fn store_file(base: &Path, src: &Path, sha: &str) -> Result<PathBuf, AgentError> {
-    if sha.len() < 2 {
-        return Err(AgentError::Other(format!("invalid content hash '{sha}'")));
-    }
-    let dir = base.join("blobs").join(&sha[0..2]);
-    fs::create_dir_all(&dir)?;
-    let dest = dir.join(sha);
-    if !dest.exists() {
-        fs::copy(src, &dest)?;
-    }
-    Ok(dest)
-}
-
-/// Extract a zip archive into `dest` (created if needed). `enclosed_name`
-/// sanitizes each entry path, so this is safe against zip-slip traversal.
-pub fn extract_zip(archive: &Path, dest: &Path) -> Result<(), AgentError> {
-    let file = File::open(archive)?;
-    let mut zip =
-        zip::ZipArchive::new(file).map_err(|e| AgentError::Other(format!("open zip: {e}")))?;
-    fs::create_dir_all(dest)?;
-    for i in 0..zip.len() {
-        let mut entry = zip
-            .by_index(i)
-            .map_err(|e| AgentError::Other(format!("read zip entry: {e}")))?;
-        // `enclosed_name` returns `None` for traversal-unsafe paths — skip them.
-        let Some(rel) = entry.enclosed_name() else {
-            continue;
-        };
-        let out = dest.join(rel);
-        if entry.is_dir() {
-            fs::create_dir_all(&out)?;
-        } else {
-            if let Some(parent) = out.parent() {
-                fs::create_dir_all(parent)?;
-            }
-            let mut outfile = File::create(&out)?;
-            io::copy(&mut entry, &mut outfile)?;
-        }
-    }
-    Ok(())
-}
-
-/// The working directory for one artifact of a target: `<base>/work/<target>/<artifact>`.
-pub fn work_dir(base: &Path, target_id: &str, artifact_id: &str) -> PathBuf {
-    base.join("work").join(target_id).join(artifact_id)
-}
-
-#[cfg(test)]
-#[allow(clippy::expect_used, clippy::unwrap_used)]
-mod tests {
-    use super::*;
-
-    /// A unique scratch directory, removed on drop.
-    struct Scratch(PathBuf);
-    impl Scratch {
-        fn new() -> Self {
-            let p = std::env::temp_dir().join(format!("cs-ingest-{}", uuid::Uuid::new_v4()));
-            fs::create_dir_all(&p).expect("mkdir scratch");
-            Self(p)
-        }
-        fn path(&self) -> &Path {
-            &self.0
-        }
-    }
-    impl Drop for Scratch {
-        fn drop(&mut self) {
-            let _ = fs::remove_dir_all(&self.0);
-        }
-    }
-
-    #[test]
-    fn hash_is_stable_and_reports_size() {
-        let dir = Scratch::new();
-        let f = dir.path().join("a.bin");
-        fs::write(&f, b"hello world").expect("write");
-        let (sha, size) = hash_file(&f).expect("hash");
-        assert_eq!(size, 11);
-        // Known SHA-256 of "hello world".
-        assert_eq!(
-            sha,
-            "b94d27b9934d3e08a52e52d7da7dabfac484efe37a5380ee9088f7ace2efcde9"
-        );
-    }
-
-    #[test]
-    fn store_is_content_addressed_and_idempotent() {
-        let base = Scratch::new();
-        let src = base.path().join("src.bin");
-        fs::write(&src, b"payload").expect("write");
-        let (sha, _) = hash_file(&src).expect("hash");
-        let p1 = store_file(base.path(), &src, &sha).expect("store");
-        let p2 = store_file(base.path(), &src, &sha).expect("store again");
-        assert_eq!(p1, p2);
-        assert!(p1.ends_with(&sha));
-        assert!(p1.starts_with(base.path().join("blobs").join(&sha[0..2])));
-        assert_eq!(fs::read(&p1).expect("read"), b"payload");
-    }
-
-    #[test]
-    fn extract_zip_writes_entries() {
-        let base = Scratch::new();
-        let archive = base.path().join("a.zip");
-        {
-            let file = File::create(&archive).expect("create");
-            let mut w = zip::ZipWriter::new(file);
-            let opts: zip::write::SimpleFileOptions = Default::default();
-            w.start_file("dir/hello.txt", opts).expect("start");
-            io::Write::write_all(&mut w, b"hi").expect("write");
-            w.finish().expect("finish");
-        }
-        let dest = base.path().join("out");
-        extract_zip(&archive, &dest).expect("extract");
-        assert_eq!(
-            fs::read_to_string(dest.join("dir/hello.txt")).expect("read"),
-            "hi"
-        );
-    }
-}
@@ -1,416 +0,0 @@
-//! Artifact ingest.
-//!
-//! Normalizes each [`Artifact`] on an [`OnboardedTarget`] into a local working
-//! path plus recorded metadata (content hash, size, discovered facts) that the
-//! classifier and scanners consume. Every blob is SHA-256 hashed — that digest
-//! is also the reconciliation key against sibling products (a firmware sha256
-//! matches tramiton's `Artifact.sha256`).
-
-pub(crate) mod blob;
-
-use std::collections::HashMap;
-use std::path::{Path, PathBuf};
-
-use compliance_core::models::{Artifact, ArtifactKind, DetectedFact, OnboardedTarget};
-use compliance_core::AgentConfig;
-
-use crate::error::AgentError;
-use crate::pipeline::git::{GitOps, RepoCredentials};
-
-/// The paths and identifiers an ingest needs. Decoupled from the full
-/// [`AgentConfig`] so ingest is testable without a complete config.
-pub struct IngestContext<'a> {
-    /// Base directory for content-addressed blobs and working dirs.
-    pub artifact_store_base: &'a Path,
-    /// Base directory for git clones.
-    pub git_clone_base: &'a str,
-    /// Default SSH key path (used when an artifact provides none).
-    pub ssh_key_path: &'a str,
-    /// The id of the target these artifacts belong to (namespaces working dirs).
-    pub target_id: &'a str,
-}
-
-impl<'a> IngestContext<'a> {
-    /// Build an ingest context from the agent config for a given target.
-    pub fn from_config(config: &'a AgentConfig, target_id: &'a str) -> Self {
-        Self {
-            artifact_store_base: Path::new(&config.artifact_store_base_path),
-            git_clone_base: &config.git_clone_base_path,
-            ssh_key_path: &config.ssh_key_path,
-            target_id,
-        }
-    }
-}
-
-/// The result of ingesting one artifact.
-pub struct IngestedArtifact {
-    /// The artifact this corresponds to ([`Artifact::id`]).
-    pub artifact_id: String,
-    /// The artifact kind.
-    pub kind: ArtifactKind,
-    /// Local working path (clone dir, extracted dir, or blob file). `None` for
-    /// artifacts with no on-disk form (live URL, plaintext, container ref).
-    pub working_path: Option<PathBuf>,
-    /// SHA-256 of the content (blobs) or git head SHA (git repos).
-    pub content_hash: Option<String>,
-    /// Stored blob size in bytes, when applicable.
-    pub size_bytes: Option<u64>,
-    /// Facts discovered during ingest.
-    pub facts: Vec<DetectedFact>,
-}
-
-/// All ingested artifacts for a target, keyed by artifact id.
-pub struct IngestSet {
-    /// The ingested artifacts, keyed by [`Artifact::id`].
-    pub by_artifact: HashMap<String, IngestedArtifact>,
-}
-
-impl IngestSet {
-    /// The working paths of every ingested artifact that has one — the input the
-    /// classifier expects.
-    pub fn working_paths(&self) -> HashMap<String, PathBuf> {
-        self.by_artifact
-            .iter()
-            .filter_map(|(id, a)| a.working_path.clone().map(|p| (id.clone(), p)))
-            .collect()
-    }
-
-    /// The ingest result for a specific artifact.
-    pub fn get(&self, artifact_id: &str) -> Option<&IngestedArtifact> {
-        self.by_artifact.get(artifact_id)
-    }
-}
-
-/// Ingest every artifact on a target.
-pub fn ingest_all(
-    target: &OnboardedTarget,
-    ctx: &IngestContext<'_>,
-) -> Result<IngestSet, AgentError> {
-    let mut by_artifact = HashMap::new();
-    for artifact in &target.artifacts {
-        let ingested = ingest_artifact(artifact, ctx)?;
-        by_artifact.insert(artifact.id.clone(), ingested);
-    }
-    Ok(IngestSet { by_artifact })
-}
-
-/// Ingest a single artifact, dispatching on its kind.
-pub fn ingest_artifact(
-    artifact: &Artifact,
-    ctx: &IngestContext<'_>,
-) -> Result<IngestedArtifact, AgentError> {
-    match artifact.kind {
-        ArtifactKind::GitRepo => ingest_git(artifact, ctx),
-        ArtifactKind::SourceArchive | ArtifactKind::MobilePackage | ArtifactKind::PlcProject => {
-            ingest_blob(artifact, ctx, true)
-        }
-        ArtifactKind::FirmwareImage => ingest_blob(artifact, ctx, false),
-        ArtifactKind::ContainerImage => Ok(metadata_only(
-            artifact,
-            DetectedFact::new("container_ref", artifact.source_ref.as_str(), "ingest"),
-        )),
-        ArtifactKind::LiveUrl => Ok(metadata_only(
-            artifact,
-            DetectedFact::new("live_url", artifact.source_ref.as_str(), "ingest"),
-        )),
-        ArtifactKind::PlaintextDescription => Ok(metadata_only(
-            artifact,
-            DetectedFact::new(
-                "description_len",
-                artifact.source_ref.len().to_string(),
-                "ingest",
-            ),
-        )),
-    }
-}
-
-/// Clone (or fetch) a git artifact, recording the head SHA as the content hash.
-fn ingest_git(
-    artifact: &Artifact,
-    ctx: &IngestContext<'_>,
-) -> Result<IngestedArtifact, AgentError> {
-    let creds = credentials_for(artifact, ctx.ssh_key_path);
-    let git_ops = GitOps::new(ctx.git_clone_base, creds);
-    let repo_path = git_ops.clone_or_fetch(&artifact.source_ref, &artifact.id)?;
-    let head = GitOps::get_head_sha(&repo_path).ok();
-    Ok(IngestedArtifact {
-        artifact_id: artifact.id.clone(),
-        kind: artifact.kind,
-        working_path: Some(repo_path),
-        content_hash: head,
-        size_bytes: None,
-        facts: Vec::new(),
-    })
-}
-
-/// Store a blob artifact content-addressed. When `extract` is set and the blob
-/// is a zip container (source archive, APK/AAB/IPA), also unpack it into a
-/// working directory; otherwise the working path is the stored blob.
-fn ingest_blob(
-    artifact: &Artifact,
-    ctx: &IngestContext<'_>,
-    extract: bool,
-) -> Result<IngestedArtifact, AgentError> {
-    let base = ctx.artifact_store_base;
-    let src = local_source(artifact)?;
-    let (sha, size) = blob::hash_file(&src)?;
-    let stored = blob::store_file(base, &src, &sha)?;
-
-    let mut facts = Vec::new();
-    let working_path = if extract {
-        let dest = blob::work_dir(base, ctx.target_id, &artifact.id);
-        match blob::extract_zip(&stored, &dest) {
-            Ok(()) => dest,
-            Err(e) => {
-                // Not a zip container — this is a single uploaded file (e.g. a
-                // `.st`/`.xml` PLC project or a `.tar.gz`). The content-addressed
-                // blob has no extension, so materialize it into a working dir
-                // under its original name; extension-based scanners (PLC) can then
-                // discover it and report a readable path.
-                facts.push(DetectedFact::new(
-                    "archive_unextracted",
-                    e.to_string(),
-                    "ingest",
-                ));
-                match materialize_single(&stored, &dest, &blob_file_name(artifact)) {
-                    Ok(dir) => dir,
-                    Err(copy_err) => {
-                        facts.push(DetectedFact::new(
-                            "materialize_failed",
-                            copy_err.to_string(),
-                            "ingest",
-                        ));
-                        stored.clone()
-                    }
-                }
-            }
-        }
-    } else {
-        stored.clone()
-    };
-
-    Ok(IngestedArtifact {
-        artifact_id: artifact.id.clone(),
-        kind: artifact.kind,
-        working_path: Some(working_path),
-        content_hash: Some(sha),
-        size_bytes: Some(size),
-        facts,
-    })
-}
-
-/// Copy a stored blob into `dest`/`name`, returning `dest`. Used when an
-/// "extractable" artifact turns out to be a single file rather than an archive.
-fn materialize_single(stored: &Path, dest: &Path, name: &str) -> Result<PathBuf, AgentError> {
-    std::fs::create_dir_all(dest)?;
-    std::fs::copy(stored, dest.join(name))?;
-    Ok(dest.to_path_buf())
-}
-
-/// A safe, single-segment file name for an artifact, preserving the original
-/// extension so scanners can identify it. Derives from `source_ref` (the
-/// uploaded/original file name); `file_name` strips any directory components,
-/// so this is traversal-safe. Falls back to the artifact id.
-fn blob_file_name(artifact: &Artifact) -> String {
-    Path::new(&artifact.source_ref)
-        .file_name()
-        .and_then(|n| n.to_str())
-        .map(str::to_string)
-        .filter(|s| !s.is_empty())
-        .unwrap_or_else(|| format!("artifact-{}", artifact.id))
-}
-
-/// An artifact with no on-disk form: record a single fact, no hash/path.
-fn metadata_only(artifact: &Artifact, fact: DetectedFact) -> IngestedArtifact {
-    IngestedArtifact {
-        artifact_id: artifact.id.clone(),
-        kind: artifact.kind,
-        working_path: None,
-        content_hash: None,
-        size_bytes: None,
-        facts: vec![fact],
-    }
-}
-
-/// The local file backing a blob artifact: its `stored_path` if already
-/// uploaded, else its `source_ref` interpreted as a filesystem path.
-fn local_source(artifact: &Artifact) -> Result<PathBuf, AgentError> {
-    let path = artifact
-        .stored_path
-        .as_deref()
-        .unwrap_or(artifact.source_ref.as_str());
-    let path = PathBuf::from(path);
-    if !path.exists() {
-        return Err(AgentError::Other(format!(
-            "artifact {} source not found at {}",
-            artifact.id,
-            path.display()
-        )));
-    }
-    Ok(path)
-}
-
-/// Build git credentials from an artifact's auth plus a default SSH key path.
-fn credentials_for(artifact: &Artifact, default_ssh_key_path: &str) -> RepoCredentials {
-    let auth = artifact.auth.as_ref();
-    RepoCredentials {
-        ssh_key_path: auth
-            .and_then(|a| a.ssh_key_path.clone())
-            .or_else(|| Some(default_ssh_key_path.to_string())),
-        auth_token: auth.and_then(|a| a.secret.clone()),
-        auth_username: auth.and_then(|a| a.username.clone()),
-    }
-}
-
-#[cfg(test)]
-#[allow(clippy::expect_used, clippy::unwrap_used)]
-mod tests {
-    use super::*;
-    use compliance_core::models::{ArtifactAuth, TargetType};
-
-    /// A unique scratch directory, removed on drop.
-    struct Scratch(PathBuf);
-    impl Scratch {
-        fn new() -> Self {
-            let p = std::env::temp_dir().join(format!("cs-ingest-mod-{}", uuid::Uuid::new_v4()));
-            std::fs::create_dir_all(&p).expect("mkdir scratch");
-            Self(p)
-        }
-    }
-    impl Drop for Scratch {
-        fn drop(&mut self) {
-            let _ = std::fs::remove_dir_all(&self.0);
-        }
-    }
-
-    fn ctx_for<'a>(store: &'a Path, target_id: &'a str) -> IngestContext<'a> {
-        IngestContext {
-            artifact_store_base: store,
-            git_clone_base: "/tmp/cs-ingest-test-repos",
-            ssh_key_path: "/tmp/cs-ingest-test-ssh",
-            target_id,
-        }
-    }
-
-    #[test]
-    fn firmware_blob_is_hashed_and_stored() {
-        let scratch = Scratch::new();
-        let store = scratch.0.join("store");
-        let fw = scratch.0.join("fw.bin");
-        std::fs::write(&fw, b"firmware-bytes").expect("write");
-
-        let ctx = ctx_for(&store, "t1");
-        let artifact = Artifact::firmware_image(fw.to_string_lossy().to_string());
-        let out = ingest_artifact(&artifact, &ctx).expect("ingest");
-
-        assert_eq!(out.kind, ArtifactKind::FirmwareImage);
-        assert_eq!(out.size_bytes, Some(14));
-        let sha = out.content_hash.expect("hash");
-        assert_eq!(sha.len(), 64);
-        // working path is the content-addressed blob
-        let wp = out.working_path.expect("working path");
-        assert!(wp.starts_with(store.join("blobs")));
-    }
-
-    #[test]
-    fn live_url_has_no_blob() {
-        let scratch = Scratch::new();
-        let store = scratch.0.join("store");
-        let ctx = ctx_for(&store, "t1");
-        let artifact = Artifact::live_url("https://example.com");
-        let out = ingest_artifact(&artifact, &ctx).expect("ingest");
-        assert!(out.working_path.is_none());
-        assert!(out.content_hash.is_none());
-        assert!(out.facts.iter().any(|f| f.key == "live_url"));
-    }
-
-    #[test]
-    fn ingest_all_collects_working_paths() {
-        let scratch = Scratch::new();
-        let store = scratch.0.join("store");
-        let fw = scratch.0.join("fw.bin");
-        std::fs::write(&fw, b"abc").expect("write");
-        let ctx = ctx_for(&store, "t1");
-
-        let mut target = OnboardedTarget::new("t".to_string(), TargetType::FirmwareBareMetal);
-        target
-            .artifacts
-            .push(Artifact::firmware_image(fw.to_string_lossy().to_string()));
-        target.artifacts.push(Artifact::live_url("https://x"));
-
-        let set = ingest_all(&target, &ctx).expect("ingest all");
-        assert_eq!(set.by_artifact.len(), 2);
-        // Only the firmware artifact yields a working path.
-        assert_eq!(set.working_paths().len(), 1);
-    }
-
-    #[test]
-    fn credentials_prefer_artifact_auth() {
-        let mut artifact = Artifact::git_repo("https://git/x", "main");
-        artifact.auth = Some(ArtifactAuth {
-            method: "token".to_string(),
-            username: Some("bob".to_string()),
-            secret: Some("pat".to_string()),
-            ..Default::default()
-        });
-        let creds = credentials_for(&artifact, "/default/ssh/key");
-        assert_eq!(creds.auth_token.as_deref(), Some("pat"));
-        assert_eq!(creds.auth_username.as_deref(), Some("bob"));
-    }
-
-    #[test]
-    fn credentials_fall_back_to_default_ssh_key() {
-        let artifact = Artifact::git_repo("git@host:x.git", "main");
-        let creds = credentials_for(&artifact, "/default/ssh/key");
-        assert_eq!(creds.ssh_key_path.as_deref(), Some("/default/ssh/key"));
-        assert!(creds.auth_token.is_none());
-    }
-
-    /// A single uploaded PLC file (not an archive) must land in a working dir
-    /// under its original name so the PLC scanner can discover it by extension
-    /// and report a readable path — the demo's upload → scan path.
-    #[test]
-    fn single_uploaded_plc_file_is_materialized_and_scannable() {
-        use compliance_core::models::PlcFormat;
-
-        let scratch = Scratch::new();
-        let store = scratch.0.join("store");
-        // Simulate the upload handler: bytes written to an `uploads/` path,
-        // `source_ref` carrying the original (clean) file name.
-        let uploads = scratch.0.join("uploads");
-        std::fs::create_dir_all(&uploads).expect("mkdir uploads");
-        let uploaded = uploads.join("a1b2c3_pump_station.st");
-        std::fs::write(
-            &uploaded,
-            "PROGRAM P\nVAR\n  ApiKey : STRING := 'sk-live-1234';\nEND_VAR\nEND_PROGRAM\n",
-        )
-        .expect("write st");
-
-        let mut artifact = Artifact::plc_project("pump_station.st", PlcFormat::StructuredText);
-        artifact.stored_path = Some(uploaded.to_string_lossy().to_string());
-
-        let ctx = ctx_for(&store, "t-plc");
-        let out = ingest_artifact(&artifact, &ctx).expect("ingest");
-
-        // Working path is a directory (not the extensionless blob) holding the
-        // file under its original name.
-        let wp = out.working_path.expect("working path");
-        assert!(wp.is_dir(), "expected a working dir, got {wp:?}");
-        assert!(wp.join("pump_station.st").is_file());
-
-        // The PLC scanner finds the hardcoded credential and reports a clean path.
-        let findings = crate::pipeline::plc::analyze_tree(&wp, "t-plc");
-        assert!(
-            !findings.is_empty(),
-            "scanner should flag the uploaded file"
-        );
-        assert!(findings
-            .iter()
-            .any(|f| f.rule_id.as_deref() == Some("plc-hardcoded-credential")));
-        assert_eq!(
-            findings[0].file_path.as_deref(),
-            Some("pump_station.st"),
-            "finding should reference the original file name"
-        );
-    }
-}
@@ -1,20 +0,0 @@
-// Library entrypoint — re-exports for integration tests and the binary.
-
-pub mod agent;
-pub mod api;
-pub mod classify;
-pub mod config;
-pub mod controls;
-pub mod database;
-pub mod error;
-pub mod ingest;
-pub mod llm;
-pub mod pentest;
-pub mod pipeline;
-pub mod rag;
-pub mod scheduler;
-pub mod ssh;
-#[allow(dead_code)]
-pub mod trackers;
-pub mod webhooks;
-pub mod werkbank;
@@ -19,17 +19,12 @@ impl LlmClient {
        model: String,
        embed_model: String,
    ) -> Self {
-        let http = reqwest::Client::builder()
-            .timeout(std::time::Duration::from_secs(300))
-            .connect_timeout(std::time::Duration::from_secs(10))
-            .build()
-            .unwrap_or_default();
        Self {
            base_url,
            api_key,
            model,
            embed_model,
-            http,
+            http: reqwest::Client::new(),
        }
    }

@@ -5,20 +5,15 @@ use compliance_core::models::Finding;
 use crate::error::AgentError;
 use crate::llm::LlmClient;

-const DESCRIPTION_SYSTEM_PROMPT: &str = r#"You are a security engineer writing a bug tracker issue for a developer to fix. Be direct and actionable — developers skim issue descriptions, so lead with what matters.
+const DESCRIPTION_SYSTEM_PROMPT: &str = r#"You are a security engineer writing issue descriptions for a bug tracker. Generate a clear, actionable issue body in Markdown format that includes:

-Format in Markdown:
+1. **Summary**: 1-2 sentence overview
+2. **Evidence**: Code location, snippet, and what was detected
+3. **Impact**: What could happen if not fixed
+4. **Remediation**: Step-by-step fix instructions
+5. **References**: Relevant CWE/CVE links if applicable

-1. **What**: 1 sentence — what's wrong and where (file:line)
-2. **Why it matters**: 1-2 sentences — concrete impact if not fixed. Avoid generic "could lead to" phrasing; describe the specific attack or failure scenario.
-3. **Fix**: The specific code change needed. Use a code block with the corrected code if possible. If the fix is configuration-based, show the exact config change.
-4. **References**: CWE/CVE link if applicable (one line, not a section)
-
-Rules:
- No filler paragraphs or background explanations
- No restating the finding title in the body
- Code blocks should show the FIX, not the vulnerable code (the developer can see that in the diff)
- If the remediation is a one-liner, just say it — don't wrap it in a section header"#;
+Keep it concise and professional. Use code blocks for code snippets."#;

 pub async fn generate_issue_description(
    llm: &Arc<LlmClient>,
@@ -22,11 +22,6 @@ struct EmbeddingData {
    index: usize,
 }

-/// Max inputs per embedding request. The bge/OpenAI-like backends cap the input
-/// array (bge-multilingual-gemma2 rejects >25 with "batch size overflow"), so we
-/// chunk larger corpora — a whole control catalog (~1.8k) would otherwise 500.
-const EMBED_BATCH_SIZE: usize = 16;
-
 // ── Embedding implementation ───────────────────────────────────

 impl LlmClient {
@@ -34,21 +29,8 @@ impl LlmClient {
        &self.embed_model
    }

-    /// Generate embeddings for a batch of texts, chunking into backend-sized
-    /// requests and preserving input order across chunks.
+    /// Generate embeddings for a batch of texts
    pub async fn embed(&self, texts: Vec<String>) -> Result<Vec<Vec<f64>>, AgentError> {
-        if texts.is_empty() {
-            return Ok(Vec::new());
-        }
-        let mut out = Vec::with_capacity(texts.len());
-        for chunk in texts.chunks(EMBED_BATCH_SIZE) {
-            out.extend(self.embed_batch(chunk.to_vec()).await?);
-        }
-        Ok(out)
-    }
-
-    /// Embed one backend-sized batch (≤ [`EMBED_BATCH_SIZE`]) in a single request.
-    async fn embed_batch(&self, texts: Vec<String>) -> Result<Vec<Vec<f64>>, AgentError> {
        let url = format!("{}/v1/embeddings", self.base_url.trim_end_matches('/'));

        let request_body = EmbeddingRequest {
@@ -90,33 +72,3 @@ impl LlmClient {
        Ok(data.into_iter().map(|d| d.embedding).collect())
    }
 }
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use secrecy::SecretString;
-
-    fn client() -> LlmClient {
-        LlmClient::new(
-            "http://unused".into(),
-            SecretString::from(String::new()),
-            "m".into(),
-            "e".into(),
-        )
-    }
-
-    #[tokio::test]
-    async fn empty_input_makes_no_request() {
-        // Must short-circuit before any HTTP call (base_url is unroutable).
-        let out = client().embed(Vec::new()).await.unwrap();
-        assert!(out.is_empty());
-    }
-
-    #[test]
-    fn batch_size_is_within_backend_cap() {
-        assert!(
-            EMBED_BATCH_SIZE <= 25,
-            "must stay under the bge 25-input cap"
-        );
-    }
-}
@@ -5,24 +5,7 @@ use compliance_core::models::Finding;
 use crate::error::AgentError;
 use crate::llm::LlmClient;

-const FIX_SYSTEM_PROMPT: &str = r#"You are a security engineer suggesting a code fix. Return ONLY the corrected code that replaces the vulnerable snippet — no explanations, no markdown fences, no before/after comparison.
-
-Rules:
- The fix must be a drop-in replacement for the vulnerable code
- Preserve the original code's style, indentation, and naming conventions
- Add at most one brief inline comment on the changed line explaining the security fix
- If the fix requires importing a new module, include the import on a separate line prefixed with the language's comment syntax + "Add import: "
- Do not refactor, rename variables, or "improve" unrelated code
- If the vulnerability is a false positive and the code is actually safe, return the original code unchanged with a comment explaining why no fix is needed
-
-Language-specific fix guidance:
- Rust: use `?` for error propagation, prefer `SecretString` for secrets, use parameterized queries with `sqlx`/`diesel`
- Python: use parameterized queries (never f-strings in SQL), use `secrets` module not `random`, use `subprocess.run([...])` list form, use `markupsafe.escape()` for HTML
- Go: use `sql.Query` with `$1`/`?` placeholders, use `crypto/rand` not `math/rand`, use `html/template` not `text/template`, return errors don't panic
- Java/Kotlin: use `PreparedStatement` with `?` params, use `SecureRandom`, use `Jsoup.clean()` for HTML sanitization, use `@Valid` for input validation
- Ruby: use ActiveRecord parameterized finders, use `SecureRandom`, use `ERB::Util.html_escape`, use `strong_parameters`
- PHP: use PDO prepared statements with `:param` or `?`, use `random_bytes()`/`random_int()`, use `htmlspecialchars()` with `ENT_QUOTES`, use `password_hash(PASSWORD_BCRYPT)`
- C/C++: use `snprintf` not `sprintf`, use bounds-checked APIs, free resources in reverse allocation order, use `memset_s` for secret cleanup"#;
+const FIX_SYSTEM_PROMPT: &str = r#"You are a security engineer. Given a security finding with code context, suggest a concrete code fix. Return ONLY the fixed code snippet that can directly replace the vulnerable code. Include brief inline comments explaining the fix."#;

 pub async fn suggest_fix(llm: &Arc<LlmClient>, finding: &Finding) -> Result<String, AgentError> {
    let user_prompt = format!(
@@ -1,138 +1,69 @@
 // System prompts for multi-pass LLM code review.
 // Each pass focuses on a different aspect to avoid overloading a single prompt.

-pub const LOGIC_REVIEW_PROMPT: &str = r#"You are a senior software engineer reviewing a code diff. Report ONLY genuine logic bugs that would cause incorrect behavior at runtime.
+pub const LOGIC_REVIEW_PROMPT: &str = r#"You are a senior software engineer reviewing code changes. Focus ONLY on logic and correctness issues.

-Report:
- Off-by-one errors, wrong comparisons, missing edge cases that cause wrong results
- Incorrect control flow that produces wrong output (not style preferences)
- Actual race conditions with concrete shared-state mutation (not theoretical ones)
- Resource leaks where cleanup is truly missing (not just "could be improved")
- Wrong variable used (copy-paste errors) — must be provably wrong, not just suspicious
- Swallowed errors that silently hide failures in a way that matters
+Look for:
+- Off-by-one errors, wrong comparisons, missing edge cases
+- Incorrect control flow (unreachable code, missing returns, wrong loop conditions)
+- Race conditions or concurrency bugs
+- Resource leaks (unclosed handles, missing cleanup)
+- Wrong variable used (copy-paste errors)
+- Incorrect error handling (swallowed errors, wrong error type)

-Do NOT report:
- Style, naming, formatting, documentation, or code organization preferences
- Theoretical issues without a concrete triggering scenario
- "Potential" problems that require assumptions not supported by the visible code
- Complexity or function length — that's a separate review pass
+Ignore: style, naming, formatting, documentation, minor improvements.

-Language-idiomatic patterns that are NOT bugs (do not flag these):
- Rust: `||`/`&&` short-circuit evaluation, variable shadowing, `let` rebinding, `clone()`, `impl` blocks, `match` arms with guards, `?` operator chaining, `unsafe` blocks with safety comments
- Python: duck typing, EAFP pattern (try/except vs check-first), `*args`/`**kwargs`, walrus operator `:=`, truthiness checks on containers, bare `except:` in top-level handlers
- Go: multiple return values for errors, `if err != nil` patterns, goroutine + channel patterns, blank identifier `_`, named returns, `defer` for cleanup, `init()` functions
- Java/Kotlin: checked exception patterns, method overloading, `Optional` vs null checks, Kotlin `?.` safe calls, `!!` non-null assertions in tests, `when` exhaustive matching, companion objects, `lateinit`
- Ruby: monkey patching in libraries, method_missing, blocks/procs/lambdas, `rescue => e` patterns, `send`/`respond_to?` metaprogramming, `nil` checks via `&.` safe navigation
- PHP: loose comparisons with `==` (only flag if `===` was clearly intended), `@` error suppression in legacy code, `isset()`/`empty()` patterns, magic methods (`__get`, `__call`), array functions as callbacks
- C/C++: RAII patterns, move semantics, `const_cast`/`static_cast` in appropriate contexts, macro usage for platform compat, pointer arithmetic in low-level code, `goto` for cleanup in C
-
-Severity guide:
- high: Will cause incorrect behavior in normal usage
- medium: Will cause incorrect behavior in edge cases
- low: Minor correctness concern with limited blast radius
-
-Prefer returning [] over reporting low-confidence guesses. A false positive wastes more developer time than a missed low-severity issue.
-
-Respond with a JSON array (no markdown fences):
+For each issue found, respond with a JSON array:
 [{"title": "...", "description": "...", "severity": "high|medium|low", "file": "...", "line": N, "suggestion": "..."}]

 If no issues found, respond with: []"#;

-pub const SECURITY_REVIEW_PROMPT: &str = r#"You are a security engineer reviewing a code diff. Report ONLY exploitable security vulnerabilities with a realistic attack scenario.
+pub const SECURITY_REVIEW_PROMPT: &str = r#"You are a security engineer reviewing code changes. Focus ONLY on security vulnerabilities.

-Report:
- Injection vulnerabilities (SQL, command, XSS, template) where untrusted input reaches a sink
- Authentication/authorization bypasses with a concrete exploit path
- Sensitive data exposure: secrets in code, credentials in logs, PII leaks
- Insecure cryptography: weak algorithms, predictable randomness, hardcoded keys
- Path traversal, SSRF, open redirects — only where user input reaches the vulnerable API
- Unsafe deserialization of untrusted data
- Missing input validation at EXTERNAL trust boundaries (user input, API responses)
+Look for:
+- Injection vulnerabilities (SQL, command, XSS, template injection)
+- Authentication/authorization bypasses
+- Sensitive data exposure (logging secrets, hardcoded credentials)
+- Insecure cryptography (weak algorithms, predictable randomness)
+- Path traversal, SSRF, open redirects
+- Unsafe deserialization
+- Missing input validation at trust boundaries

-Do NOT report:
- Internal code that only handles trusted/validated data
- Hash functions used for non-security purposes (dedup fingerprints, cache keys, content addressing)
- Logging of non-sensitive operational data (finding titles, counts, performance metrics)
- "Information disclosure" for data that is already public or user-facing
- Code style, performance, or general quality issues
- Missing validation on internal function parameters (trust the caller within the same module/crate/package)
- Theoretical attacks that require preconditions not present in the code
+Ignore: code style, performance, general quality.

-Language-specific patterns that are NOT vulnerabilities (do not flag these):
- Python: `pickle` used on trusted internal data, `eval()`/`exec()` on hardcoded strings, `subprocess` with hardcoded commands, Django `mark_safe()` on static content, `assert` in non-security contexts
- Go: `crypto/rand` is secure (don't confuse with `math/rand`), `sql.DB` with parameterized queries is safe, `http.ListenAndServe` without TLS in dev/internal, error strings in responses (Go convention)
- Java/Kotlin: Spring Security annotations are sufficient auth checks, `@Transactional` provides atomicity, JPA parameterized queries are safe, Kotlin `require()`/`check()` are assertion patterns not vulnerabilities
- Ruby: Rails `params.permit()` is input validation, `render html:` with `html_safe` on generated content, ActiveRecord parameterized finders are safe, Devise/Warden patterns for auth
- PHP: PDO prepared statements are safe, Laravel Eloquent is parameterized, `htmlspecialchars()` is XSS mitigation, Symfony security voters are auth checks, `password_hash()`/`password_verify()` are correct bcrypt usage
- C/C++: `strncpy`/`snprintf` are bounds-checked (vs `strcpy`/`sprintf`), smart pointers manage memory, RAII handles cleanup, `static_assert` is compile-time only, OpenSSL with proper context setup
- Rust: `sha2`/`blake3` for fingerprinting is not "weak crypto", `unsafe` with documented invariants, `secrecy::SecretString` properly handles secrets
-
-Severity guide:
- critical: Remote code execution, auth bypass, or data breach with no preconditions
- high: Exploitable vulnerability requiring minimal preconditions
- medium: Vulnerability requiring specific conditions or limited impact
-
-Prefer returning [] over reporting speculative vulnerabilities. Every false positive erodes trust in the scanner.
-
-Respond with a JSON array (no markdown fences):
+For each issue found, respond with a JSON array:
 [{"title": "...", "description": "...", "severity": "critical|high|medium", "file": "...", "line": N, "cwe": "CWE-XXX", "suggestion": "..."}]

 If no issues found, respond with: []"#;

-pub const CONVENTION_REVIEW_PROMPT: &str = r#"You are a code reviewer checking for convention violations that indicate likely bugs. Report ONLY deviations from the project's visible patterns that could cause real problems.
+pub const CONVENTION_REVIEW_PROMPT: &str = r#"You are a code reviewer checking adherence to project conventions. Focus ONLY on patterns that indicate likely bugs or maintenance problems.

-Report:
- Inconsistent error handling within the same module where the inconsistency could hide failures
- Public API that breaks the module's established contract (not just different style)
- Anti-patterns that are bugs in this language: e.g. `unwrap()` in Rust library code where the CI enforces `clippy::unwrap_used`, `any` defeating TypeScript's type system
+Look for:
+- Inconsistent error handling patterns within the same module
+- Public API that doesn't follow the project's established patterns
+- Missing or incorrect type annotations that could cause runtime issues
+- Anti-patterns specific to the language (e.g. unwrap in Rust library code, any in TypeScript)

-Do NOT report:
- Style preferences, formatting, naming conventions, or documentation
- Code organization suggestions ("this function should be split")
- Patterns that are valid in the language even if you'd write them differently
- "Missing type annotations" unless the code literally won't compile or causes a type inference bug
+Do NOT report: minor style preferences, documentation gaps, formatting.
+Only report issues with HIGH confidence that they deviate from the visible codebase conventions.

-Language-specific patterns that are conventional (do not flag these):
- Rust: variable shadowing, `||`/`&&` short-circuit, `let` rebinding, builder patterns, `clone()`, `From`/`Into` impl chains, `#[allow(...)]` attributes
- Python: `**kwargs` forwarding, `@property` setters, `__dunder__` methods, list comprehensions with conditions, `if TYPE_CHECKING` imports, `noqa` comments
- Go: stuttering names (`http.HTTPClient`) discouraged but not a bug, `context.Context` as first param, init() functions, `//nolint` directives, returning concrete types vs interfaces in internal code
- Java/Kotlin: builder pattern boilerplate, Lombok annotations (`@Data`, `@Builder`), Kotlin data classes, `companion object` factories, `@Suppress` annotations, checked exception wrapping
- Ruby: `attr_accessor` usage, `Enumerable` mixin patterns, `module_function`, `class << self` syntax, DSL blocks (Rake, RSpec, Sinatra routes)
- PHP: `__construct` with property promotion, Laravel facades, static factory methods, nullable types with `?`, attribute syntax `#[...]`
- C/C++: header guards vs `#pragma once`, forward declarations, `const` correctness patterns, template specialization, `auto` type deduction
-
-Severity guide:
- medium: Convention violation that will likely cause a bug or maintenance problem
- low: Convention violation that is a minor concern
-
-Return at most 3 findings. Prefer [] over marginal findings.
-
-Respond with a JSON array (no markdown fences):
+For each issue found, respond with a JSON array:
 [{"title": "...", "description": "...", "severity": "medium|low", "file": "...", "line": N, "suggestion": "..."}]

 If no issues found, respond with: []"#;

-pub const COMPLEXITY_REVIEW_PROMPT: &str = r#"You are reviewing code changes for complexity that is likely to cause bugs. Report ONLY complexity that makes the code demonstrably harder to reason about.
+pub const COMPLEXITY_REVIEW_PROMPT: &str = r#"You are reviewing code changes for excessive complexity that could lead to bugs.

-Report:
- Functions over 80 lines with multiple interleaved responsibilities (not just long)
- Deeply nested control flow (5+ levels) where flattening would prevent bugs
- Complex boolean expressions that a reader would likely misinterpret
+Look for:
+- Functions over 50 lines that should be decomposed
+- Deeply nested control flow (4+ levels)
+- Complex boolean expressions that are hard to reason about
+- Functions with 5+ parameters
+- Code duplication within the changed files

-Do NOT report:
- Functions that are long but linear and easy to follow
- Acceptable complexity: configuration setup, CLI parsing, test helpers, builder patterns
- Code that is complex because the problem is complex — only report if restructuring would reduce bug risk
- "This function does multiple things" unless you can identify a specific bug risk from the coupling
- Suggestions that would just move complexity elsewhere without reducing it
+Only report complexity issues that are HIGH risk for future bugs. Ignore acceptable complexity in configuration, CLI argument parsing, or generated code.

-Severity guide:
- medium: Complexity that has a concrete risk of causing bugs during future changes
- low: Complexity that makes review harder but is unlikely to cause bugs
-
-Return at most 2 findings. Prefer [] over reporting complexity that is justified.
-
-Respond with a JSON array (no markdown fences):
+For each issue found, respond with a JSON array:
 [{"title": "...", "description": "...", "severity": "medium|low", "file": "...", "line": N, "suggestion": "..."}]

 If no issues found, respond with: []"#;
@@ -8,46 +8,22 @@ use crate::pipeline::orchestrator::GraphContext;
 /// Maximum number of findings to include in a single LLM triage call.
 const TRIAGE_CHUNK_SIZE: usize = 30;

-const TRIAGE_SYSTEM_PROMPT: &str = r#"You are a pragmatic security triage expert. Your job is to filter out noise and keep only findings that a developer should actually fix. Be aggressive about dismissing false positives — a clean, high-signal list is more valuable than a comprehensive one.
+const TRIAGE_SYSTEM_PROMPT: &str = r#"You are a security finding triage expert. Analyze each of the following security findings with its code context and determine the appropriate action.

 Actions:
- "confirm": True positive with real impact. Keep severity as-is.
- "downgrade": Real issue but over-reported severity. Lower it.
- "upgrade": Under-reported — higher severity warranted.
- "dismiss": False positive, not exploitable, or not actionable. Remove it.
+- "confirm": The finding is a true positive at the reported severity. Keep as-is.
+- "downgrade": The finding is real but over-reported. Lower severity recommended.
+- "upgrade": The finding is under-reported. Higher severity recommended.
+- "dismiss": The finding is a false positive. Should be removed.

-Dismiss when:
- The scanner flagged a language idiom as a bug (see examples below)
- The finding is in test/example/generated/vendored code
- The "vulnerability" requires preconditions that don't exist in the code
- The finding is about code style, complexity, or theoretical concerns rather than actual bugs
- A hash function is used for non-security purposes (dedup, caching, content addressing)
- Internal logging of non-sensitive operational data is flagged as "information disclosure"
- The finding duplicates another finding already in the list
- Framework-provided security is already in place (e.g. ORM parameterized queries, CSRF middleware, auth decorators)
+Consider:
+- Is the code in a test, example, or generated file? (lower confidence for test code)
+- Does the surrounding code context confirm or refute the finding?
+- Is the finding actionable by a developer?
+- Would a real attacker be able to exploit this?

-Common false positive patterns by language (dismiss these):
- Rust: short-circuit `||`/`&&`, variable shadowing, `clone()`, `unsafe` with safety docs, `sha2` for fingerprinting
- Python: EAFP try/except, `subprocess` with hardcoded args, `pickle` on trusted data, Django `mark_safe` on static content
- Go: `if err != nil` is not "swallowed error", `crypto/rand` is secure, returning errors is not "information disclosure"
- Java/Kotlin: Spring Security annotations are valid auth, JPA parameterized queries are safe, Kotlin `!!` in tests is fine
- Ruby: Rails `params.permit` is validation, ActiveRecord finders are parameterized, `html_safe` on generated content
- PHP: PDO prepared statements are safe, Laravel Eloquent is parameterized, `htmlspecialchars` is XSS mitigation
- C/C++: `strncpy`/`snprintf` are bounds-checked, smart pointers manage memory, RAII handles cleanup
-
-Confirm only when:
- You can describe a concrete scenario where the bug manifests or the vulnerability is exploitable
- The fix is actionable (developer can change specific code to resolve it)
- The finding is in production code that handles external input or sensitive data
-
-Confidence scoring (0-10):
- 8-10: Certain true positive with clear exploit/bug scenario
- 5-7: Likely true positive, some assumptions required
- 3-4: Uncertain, needs manual review
- 0-2: Almost certainly a false positive
-
-Respond with a JSON array, one entry per finding in the same order presented (no markdown fences):
-[{"id": "<fingerprint>", "action": "confirm|downgrade|upgrade|dismiss", "confidence": 0-10, "rationale": "1-2 sentences", "remediation": "optional fix"}, ...]"#;
+Respond with a JSON array, one entry per finding in the same order they were presented:
+[{"id": "<fingerprint>", "action": "confirm|downgrade|upgrade|dismiss", "confidence": 0-10, "rationale": "brief explanation", "remediation": "optional fix suggestion"}, ...]"#;

 pub async fn triage_findings(
    llm: &Arc<LlmClient>,
@@ -1,10 +1,23 @@
-use compliance_agent::{agent, api, config, database, scheduler, ssh, webhooks};
+mod agent;
+mod api;
+pub(crate) mod config;
+mod database;
+mod error;
+mod llm;
+mod pentest;
+mod pipeline;
+mod rag;
+mod scheduler;
+mod ssh;
+#[allow(dead_code)]
+mod trackers;
+mod webhooks;

 #[tokio::main]
 async fn main() -> Result<(), Box<dyn std::error::Error>> {
    match dotenvy::dotenv() {
        Ok(path) => eprintln!("[dotenv] Loaded from: {}", path.display()),
-        Err(_) => eprintln!("[dotenv] No .env file found, using environment variables"),
+        Err(e) => eprintln!("[dotenv] FAILED: {e}"),
    }

    let _telemetry_guard = compliance_core::telemetry::init_telemetry("compliance-agent");
@@ -25,13 +38,10 @@ async fn main() -> Result<(), Box<dyn std::error::Error>> {
    }

    tracing::info!("Connecting to MongoDB...");
-    // Per-tenant pool only — the agent has no shared "default" database
-    // after M7.2-D. `mongodb_database` is now the db-name prefix used
-    // for tenant databases (`<prefix>_<tenant_id>`).
-    let db_pool =
-        database::DatabasePool::connect(&config.mongodb_uri, &config.mongodb_database).await?;
+    let db = database::Database::connect(&config.mongodb_uri, &config.mongodb_database).await?;
+    db.ensure_indexes().await?;

-    let agent = agent::ComplianceAgent::new(config.clone(), db_pool);
+    let agent = agent::ComplianceAgent::new(config.clone(), db.clone());

    tracing::info!("Starting scheduler...");
    let scheduler_agent = agent.clone();
@@ -328,7 +328,6 @@ mod tests {
            scan_schedule: String::new(),
            cve_monitor_schedule: String::new(),
            git_clone_base_path: String::new(),
-            artifact_store_base_path: String::new(),
            ssh_key_path: String::new(),
            keycloak_url: None,
            keycloak_realm: None,
@@ -340,11 +339,6 @@ mod tests {
            pentest_imap_tls: true,
            pentest_imap_username: None,
            pentest_imap_password: None,
-            admin_api_token: None,
-            tenant_registry_url: None,
-            plc_runtime: compliance_core::PlcRuntimeConfig::default(),
-            werkbank_runner_token: None,
-            breakpilot: compliance_core::config::BreakpilotConfig::default(),
        }
    }

@@ -321,38 +321,9 @@ impl PentestOrchestrator {
                                    total_findings += findings_count;

                                    let mut finding_ids: Vec<String> = Vec::new();
-                                    // Dedup findings within this tool result before inserting
-                                    let deduped_findings =
-                                        crate::pipeline::dedup::dedup_dast_findings(
-                                            result.findings,
-                                        );
-                                    for mut finding in deduped_findings {
+                                    for mut finding in result.findings {
                                        finding.scan_run_id = session_id.clone();
                                        finding.session_id = Some(session_id.clone());
-
-                                        // Check for existing duplicate in this session
-                                        let fp = crate::pipeline::dedup::compute_dast_fingerprint(
-                                            &finding,
-                                        );
-                                        let existing = self
-                                            .db
-                                            .dast_findings()
-                                            .find_one(doc! {
-                                                "session_id": &session_id,
-                                                "title": &finding.title,
-                                                "endpoint": &finding.endpoint,
-                                                "method": &finding.method,
-                                            })
-                                            .await;
-                                        if matches!(existing, Ok(Some(_))) {
-                                            tracing::debug!(
-                                                "Skipping duplicate DAST finding: {} (fp={:.12})",
-                                                finding.title,
-                                                fp,
-                                            );
-                                            continue;
-                                        }
-
                                        let insert_result =
                                            self.db.dast_findings().insert_one(&finding).await;
                                        if let Ok(res) = &insert_result {
@@ -314,21 +314,6 @@ impl PentestOrchestrator {
 - For SPA apps: a 200 HTTP status does NOT mean the page is accessible — check the actual
  page content with the browser tool to verify if it shows real data or a login redirect.

-## Finding Quality Rules
- **Do not report the same issue twice.** If multiple tools detect the same missing header or
-  vulnerability on the same endpoint, report it ONCE with the most specific tool's output.
-  For example, if the recon tool and the header scanner both find missing HSTS, report it only
-  from the header scanner (more specific).
- **Group related findings.** Missing security headers on the same endpoint are ONE finding
-  ("Missing security headers") listing all missing headers, not separate findings per header.
- **Severity must match real impact:**
-  - critical/high: Exploitable vulnerability (you can demonstrate the exploit)
-  - medium: Real misconfiguration with security implications but not directly exploitable
-  - low: Best-practice recommendation, defense-in-depth, or informational
- **Missing headers are medium at most** unless you can demonstrate a concrete exploit enabled
-  by the missing header (e.g., missing CSP + confirmed XSS = high for CSP finding).
- Console.log in third-party/vendored JS (node_modules, minified libraries) is informational only.
-
 ## Important
 - This is an authorized penetration test. All testing is permitted within the target scope.
 - Respect the rate limit of {rate_limit} requests per second.
@@ -204,202 +204,6 @@ impl CveScanner {

        Ok(results)
    }
-
-    /// Match the CODESYS **runtime** component against NVD by CPE.
-    ///
-    /// CODESYS advisories (the CoDe16 cluster and friends) are indexed in NVD by
-    /// CPE (`cpe:2.3:a:codesys:control*`) keyed off the *runtime* version — not by
-    /// the internal `Cmp*`/`Sys*` library names OSV-by-purl would look up. So we
-    /// find the runtime SBOM entry, pull every `cpe:2.3:a:codesys:*` CVE from NVD,
-    /// and keep the ones whose affected-version range covers our runtime version.
-    /// Best-effort: returns empty without an NVD key, on a network error, or when
-    /// no CODESYS runtime component is present.
-    pub async fn scan_codesys(&self, repo_id: &str, entries: &mut [SbomEntry]) -> Vec<CveAlert> {
-        let Some((name, version)) = codesys_runtime(entries) else {
-            return Vec::new();
-        };
-        let url = "https://services.nvd.nist.gov/rest/json/cves/2.0\
-                   ?virtualMatchString=cpe:2.3:a:codesys";
-        let mut req = self.http.get(url);
-        if let Some(key) = &self.nvd_api_key {
-            req = req.header("apiKey", key.as_str());
-        }
-        let body: serde_json::Value = match req.send().await {
-            Ok(r) if r.status().is_success() => match r.json().await {
-                Ok(b) => b,
-                Err(e) => {
-                    tracing::warn!("CODESYS NVD parse failed: {e}");
-                    return Vec::new();
-                }
-            },
-            Ok(r) => {
-                tracing::warn!("CODESYS NVD returned {}", r.status());
-                return Vec::new();
-            }
-            Err(e) => {
-                tracing::warn!("CODESYS NVD request failed: {e}");
-                return Vec::new();
-            }
-        };
-
-        let matched = parse_codesys_nvd(&body, &version);
-        let mut alerts = Vec::new();
-        for cve in matched {
-            if let Some(e) = entries
-                .iter_mut()
-                .find(|e| e.name == name && e.version == version)
-            {
-                e.known_vulnerabilities.push(VulnRef {
-                    id: cve.id.clone(),
-                    source: "nvd".to_string(),
-                    severity: None,
-                    url: Some(format!("https://nvd.nist.gov/vuln/detail/{}", cve.id)),
-                });
-            }
-            let mut alert = CveAlert::new(
-                cve.id,
-                repo_id.to_string(),
-                name.clone(),
-                version.clone(),
-                CveSource::Nvd,
-            );
-            alert.summary = cve.summary;
-            alert.cvss_score = cve.cvss;
-            alerts.push(alert);
-        }
-        tracing::info!(runtime = %name, version = %version, cves = alerts.len(), "CODESYS CVE match");
-        alerts
-    }
-}
-
-/// The CODESYS runtime component (name + version) from an SBOM, if present. The
-/// runtime carries the version CODESYS advisories key off; the internal library
-/// components do not.
-fn codesys_runtime(entries: &[SbomEntry]) -> Option<(String, String)> {
-    entries
-        .iter()
-        .find(|e| e.package_manager == "codesys" && e.name.starts_with("CODESYS Control"))
-        .map(|e| (e.name.clone(), e.version.clone()))
-}
-
-/// A parsed NVD CVE that affects the CODESYS runtime.
-struct CodesysCve {
-    id: String,
-    summary: Option<String>,
-    cvss: Option<f64>,
-}
-
-/// Version constraints from an NVD `cpeMatch` node.
-#[derive(Default)]
-struct CpeRange {
-    exact: Option<String>,
-    start_incl: Option<String>,
-    start_excl: Option<String>,
-    end_incl: Option<String>,
-    end_excl: Option<String>,
-}
-
-/// Parse an NVD CVE-list response and keep the CVEs whose CODESYS CPE match covers
-/// `runtime_version`.
-fn parse_codesys_nvd(body: &serde_json::Value, runtime_version: &str) -> Vec<CodesysCve> {
-    let mut out = Vec::new();
-    let Some(vulns) = body["vulnerabilities"].as_array() else {
-        return out;
-    };
-    for v in vulns {
-        let cve = &v["cve"];
-        let Some(id) = cve["id"].as_str() else {
-            continue;
-        };
-        let covered = cve["configurations"]
-            .as_array()
-            .into_iter()
-            .flatten()
-            .flat_map(|c| c["nodes"].as_array().into_iter().flatten())
-            .flat_map(|n| n["cpeMatch"].as_array().into_iter().flatten())
-            .any(|cm| {
-                cm["vulnerable"].as_bool() == Some(true)
-                    && cm["criteria"]
-                        .as_str()
-                        .is_some_and(|c| c.contains(":codesys:"))
-                    && version_matches(runtime_version, &cpe_range(cm))
-            });
-        if covered {
-            let summary = cve["descriptions"]
-                .as_array()
-                .and_then(|d| d.iter().find(|x| x["lang"].as_str() == Some("en")))
-                .and_then(|x| x["value"].as_str())
-                .map(String::from);
-            let cvss = cve["metrics"]["cvssMetricV31"]
-                .as_array()
-                .and_then(|m| m.first())
-                .and_then(|m| m["cvssData"]["baseScore"].as_f64());
-            out.push(CodesysCve {
-                id: id.to_string(),
-                summary,
-                cvss,
-            });
-        }
-    }
-    out
-}
-
-/// Build a [`CpeRange`] from an NVD `cpeMatch` object.
-fn cpe_range(cm: &serde_json::Value) -> CpeRange {
-    let exact = cm["criteria"]
-        .as_str()
-        .and_then(cpe_version)
-        .filter(|v| v != "*" && v != "-" && !v.is_empty());
-    CpeRange {
-        exact,
-        start_incl: cm["versionStartIncluding"].as_str().map(String::from),
-        start_excl: cm["versionStartExcluding"].as_str().map(String::from),
-        end_incl: cm["versionEndIncluding"].as_str().map(String::from),
-        end_excl: cm["versionEndExcluding"].as_str().map(String::from),
-    }
-}
-
-/// The version field (6th component) of a CPE 2.3 string.
-fn cpe_version(criteria: &str) -> Option<String> {
-    criteria.split(':').nth(5).map(String::from)
-}
-
-/// Whether `v` satisfies a CPE version range.
-fn version_matches(v: &str, r: &CpeRange) -> bool {
-    use std::cmp::Ordering::{Equal, Greater, Less};
-    if let Some(exact) = &r.exact {
-        return cmp_dotted(v, exact) == Equal;
-    }
-    let mut ok = true;
-    if let Some(s) = &r.start_incl {
-        ok &= cmp_dotted(v, s) != Less;
-    }
-    if let Some(s) = &r.start_excl {
-        ok &= cmp_dotted(v, s) == Greater;
-    }
-    if let Some(e) = &r.end_incl {
-        ok &= cmp_dotted(v, e) != Greater;
-    }
-    if let Some(e) = &r.end_excl {
-        ok &= cmp_dotted(v, e) == Less;
-    }
-    ok
-}
-
-/// Compare two dotted numeric versions (`4.17.0.0` vs `4.9.0.0`); missing
-/// components count as 0, non-numeric components as 0.
-fn cmp_dotted(a: &str, b: &str) -> std::cmp::Ordering {
-    let pa: Vec<u64> = a.split('.').map(|x| x.parse().unwrap_or(0)).collect();
-    let pb: Vec<u64> = b.split('.').map(|x| x.parse().unwrap_or(0)).collect();
-    for i in 0..pa.len().max(pb.len()) {
-        let x = pa.get(i).copied().unwrap_or(0);
-        let y = pb.get(i).copied().unwrap_or(0);
-        match x.cmp(&y) {
-            std::cmp::Ordering::Equal => continue,
-            other => return other,
-        }
-    }
-    std::cmp::Ordering::Equal
 }

 #[derive(serde::Deserialize)]
@@ -424,90 +228,3 @@ struct OsvVuln {
    summary: Option<String>,
    severity: Option<String>,
 }
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use std::cmp::Ordering::{Equal, Greater, Less};
-
-    fn entry(name: &str, ver: &str, pm: &str) -> SbomEntry {
-        SbomEntry::new("t".into(), name.into(), ver.into(), pm.into())
-    }
-
-    #[test]
-    fn finds_the_codesys_runtime_component() {
-        let entries = vec![
-            entry("Standard", "3.5.18.0", "codesys"),
-            entry("CODESYS Control for Linux ARM SL", "4.17.0.0", "codesys"),
-        ];
-        assert_eq!(
-            codesys_runtime(&entries),
-            Some(("CODESYS Control for Linux ARM SL".into(), "4.17.0.0".into()))
-        );
-        // Internal library components are not the runtime.
-        assert!(codesys_runtime(&[entry("Util", "3.5.21.0", "codesys")]).is_none());
-    }
-
-    #[test]
-    fn dotted_version_comparison() {
-        assert_eq!(cmp_dotted("4.17.0.0", "4.9.0.0"), Greater);
-        assert_eq!(cmp_dotted("4.9.0.0", "4.17.0.0"), Less);
-        assert_eq!(cmp_dotted("3.5.18.0", "3.5.18.0"), Equal);
-        assert_eq!(cmp_dotted("4.2", "4.2.0.0"), Equal); // missing components = 0
-    }
-
-    #[test]
-    fn version_range_matching() {
-        let end_excl = CpeRange {
-            end_excl: Some("4.9.0.0".into()),
-            ..Default::default()
-        };
-        assert!(!version_matches("4.17.0.0", &end_excl)); // patched
-        assert!(version_matches("4.5.0.0", &end_excl)); // affected
-
-        let exact = CpeRange {
-            exact: Some("3.5.16.0".into()),
-            ..Default::default()
-        };
-        assert!(version_matches("3.5.16.0", &exact));
-        assert!(!version_matches("3.5.17.0", &exact));
-
-        let span = CpeRange {
-            start_incl: Some("3.0.0.0".into()),
-            end_incl: Some("3.5.16.0".into()),
-            ..Default::default()
-        };
-        assert!(version_matches("3.5.16.0", &span));
-        assert!(!version_matches("3.5.17.0", &span));
-    }
-
-    #[test]
-    fn parses_nvd_and_matches_by_runtime_version() {
-        // Two CODESYS CVEs: one affects < 4.9 (our 4.17 is patched), one affects
-        // <= 4.20 (our 4.17 is affected). Only the latter should match.
-        let body = serde_json::json!({
-          "vulnerabilities": [
-            {"cve": {"id":"CVE-2023-0001",
-              "descriptions":[{"lang":"en","value":"old CmpBlkDrvTcp bug"}],
-              "metrics":{"cvssMetricV31":[{"cvssData":{"baseScore":7.5}}]},
-              "configurations":[{"nodes":[{"cpeMatch":[
-                {"vulnerable":true,
-                 "criteria":"cpe:2.3:a:codesys:control_for_linux_sl:*:*:*:*:*:*:*:*",
-                 "versionEndExcluding":"4.9.0.0"}
-              ]}]}]}},
-            {"cve": {"id":"CVE-2024-0002",
-              "descriptions":[{"lang":"en","value":"recent runtime bug"}],
-              "metrics":{"cvssMetricV31":[{"cvssData":{"baseScore":9.8}}]},
-              "configurations":[{"nodes":[{"cpeMatch":[
-                {"vulnerable":true,
-                 "criteria":"cpe:2.3:a:codesys:control_for_linux_sl:*:*:*:*:*:*:*:*",
-                 "versionEndIncluding":"4.20.0.0"}
-              ]}]}]}}
-          ]
-        });
-        let matched = parse_codesys_nvd(&body, "4.17.0.0");
-        let ids: Vec<&str> = matched.iter().map(|c| c.id.as_str()).collect();
-        assert_eq!(ids, vec!["CVE-2024-0002"]);
-        assert_eq!(matched[0].cvss, Some(9.8));
-    }
-}
@@ -1,7 +1,5 @@
 use sha2::{Digest, Sha256};

-use compliance_core::models::dast::DastFinding;
-
 pub fn compute_fingerprint(parts: &[&str]) -> String {
    let mut hasher = Sha256::new();
    for part in parts {
@@ -11,209 +9,9 @@ pub fn compute_fingerprint(parts: &[&str]) -> String {
    hex::encode(hasher.finalize())
 }

-/// Compute a dedup fingerprint for a DAST finding.
-///
-/// The key is derived from the *canonicalized* title (lowercased, domain names
-/// stripped, known synonyms resolved), endpoint, and HTTP method.  This lets us
-/// detect both exact duplicates (same tool reporting twice across passes) and
-/// semantic duplicates (e.g., `security_header_missing` "Missing HSTS header"
-/// vs `tls_misconfiguration` "Missing strict-transport-security header").
-pub fn compute_dast_fingerprint(f: &DastFinding) -> String {
-    let canon = canonicalize_dast_title(&f.title);
-    let endpoint = f.endpoint.to_lowercase().trim_end_matches('/').to_string();
-    let method = f.method.to_uppercase();
-    let param = f.parameter.as_deref().unwrap_or("");
-    compute_fingerprint(&[&canon, &endpoint, &method, param])
-}
-
-/// Canonicalize a DAST finding title for dedup purposes.
-///
-/// 1. Lowercase
-/// 2. Strip domain names / URLs (e.g. "for comp-dev.meghsakha.com")
-/// 3. Resolve known header synonyms (hsts ↔ strict-transport-security, etc.)
-/// 4. Strip extra whitespace
-fn canonicalize_dast_title(title: &str) -> String {
-    let mut s = title.to_lowercase();
-
-    // Strip "for <domain>" or "on <domain>" suffixes
-    // Pattern: "for <word.word...>" or "on <method> <url>"
-    if let Some(idx) = s.find(" for ") {
-        // Check if what follows looks like a domain or URL
-        let rest = &s[idx + 5..];
-        if rest.contains('.') || rest.starts_with("http") {
-            s.truncate(idx);
-        }
-    }
-    if let Some(idx) = s.find(" on ") {
-        let rest = &s[idx + 4..];
-        if rest.contains("http") || rest.contains('/') {
-            s.truncate(idx);
-        }
-    }
-
-    // Resolve known header synonyms
-    let synonyms: &[(&str, &str)] = &[
-        ("hsts", "strict-transport-security"),
-        ("csp", "content-security-policy"),
-        ("cors", "cross-origin-resource-sharing"),
-        ("xfo", "x-frame-options"),
-    ];
-    for &(short, canonical) in synonyms {
-        // Only replace whole words — check boundaries
-        if let Some(pos) = s.find(short) {
-            let before_ok = pos == 0 || !s.as_bytes()[pos - 1].is_ascii_alphanumeric();
-            let after_ok = pos + short.len() >= s.len()
-                || !s.as_bytes()[pos + short.len()].is_ascii_alphanumeric();
-            if before_ok && after_ok {
-                s = format!("{}{}{}", &s[..pos], canonical, &s[pos + short.len()..]);
-            }
-        }
-    }
-
-    // Collapse whitespace
-    s.split_whitespace().collect::<Vec<_>>().join(" ")
-}
-
-/// Deduplicate a list of DAST findings, merging evidence from duplicates.
-///
-/// Two-phase approach:
-/// 1. **Exact dedup** — group by canonicalized `(title, endpoint, method, parameter)`.
-///    Merge evidence arrays, keep the highest severity, preserve exploitable flag.
-/// 2. **CWE-based dedup** — within the same `(cwe, endpoint, method)` group, merge
-///    findings whose canonicalized titles resolve to the same subject (e.g., HSTS
-///    reported as both `security_header_missing` and `tls_misconfiguration`).
-pub fn dedup_dast_findings(findings: Vec<DastFinding>) -> Vec<DastFinding> {
-    use std::collections::HashMap;
-
-    if findings.len() <= 1 {
-        return findings;
-    }
-
-    // Phase 1: exact fingerprint dedup
-    let mut seen: HashMap<String, usize> = HashMap::new();
-    let mut deduped: Vec<DastFinding> = Vec::new();
-
-    for finding in findings {
-        let fp = compute_dast_fingerprint(&finding);
-
-        if let Some(&idx) = seen.get(&fp) {
-            // Merge into existing
-            merge_dast_finding(&mut deduped[idx], &finding);
-        } else {
-            seen.insert(fp, deduped.len());
-            deduped.push(finding);
-        }
-    }
-
-    let before = deduped.len();
-
-    // Phase 2: CWE-based related dedup
-    // Group by (cwe, endpoint_normalized, method) — only when CWE is present
-    let mut cwe_groups: HashMap<String, Vec<usize>> = HashMap::new();
-    for (i, f) in deduped.iter().enumerate() {
-        if let Some(ref cwe) = f.cwe {
-            let key = format!(
-                "{}|{}|{}",
-                cwe,
-                f.endpoint.to_lowercase().trim_end_matches('/'),
-                f.method.to_uppercase(),
-            );
-            cwe_groups.entry(key).or_default().push(i);
-        }
-    }
-
-    // For each CWE group with multiple findings, keep the one with highest severity
-    // and most evidence, merge the rest into it
-    let mut merge_map: HashMap<usize, Vec<usize>> = HashMap::new();
-    let mut remove_indices: Vec<usize> = Vec::new();
-
-    for indices in cwe_groups.values() {
-        if indices.len() <= 1 {
-            continue;
-        }
-        // Find the "primary" finding: highest severity, then most evidence, then longest description
-        let Some(&primary_idx) = indices.iter().max_by(|&&a, &&b| {
-            deduped[a]
-                .severity
-                .cmp(&deduped[b].severity)
-                .then_with(|| deduped[a].evidence.len().cmp(&deduped[b].evidence.len()))
-                .then_with(|| {
-                    deduped[a]
-                        .description
-                        .len()
-                        .cmp(&deduped[b].description.len())
-                })
-        }) else {
-            continue;
-        };
-
-        for &idx in indices {
-            if idx != primary_idx {
-                remove_indices.push(idx);
-                merge_map.entry(primary_idx).or_default().push(idx);
-            }
-        }
-    }
-
-    if !remove_indices.is_empty() {
-        remove_indices.sort_unstable();
-        remove_indices.dedup();
-
-        // Merge evidence
-        for (&primary, secondaries) in &merge_map {
-            let extra_evidence: Vec<_> = secondaries
-                .iter()
-                .flat_map(|&i| deduped[i].evidence.clone())
-                .collect();
-            let any_exploitable = secondaries.iter().any(|&i| deduped[i].exploitable);
-
-            deduped[primary].evidence.extend(extra_evidence);
-            if any_exploitable {
-                deduped[primary].exploitable = true;
-            }
-        }
-
-        // Remove merged findings (iterate in reverse to preserve indices)
-        for &idx in remove_indices.iter().rev() {
-            deduped.remove(idx);
-        }
-    }
-
-    let after = deduped.len();
-    if before != after {
-        tracing::debug!(
-            "DAST CWE-based dedup: {before} → {after} findings ({} merged)",
-            before - after
-        );
-    }
-
-    deduped
-}
-
-/// Merge a duplicate DAST finding into a primary one.
-fn merge_dast_finding(primary: &mut DastFinding, duplicate: &DastFinding) {
-    primary.evidence.extend(duplicate.evidence.clone());
-    if duplicate.severity > primary.severity {
-        primary.severity = duplicate.severity.clone();
-    }
-    if duplicate.exploitable {
-        primary.exploitable = true;
-    }
-    // Keep the longer/better description
-    if duplicate.description.len() > primary.description.len() {
-        primary.description.clone_from(&duplicate.description);
-    }
-    // Keep remediation if primary doesn't have one
-    if primary.remediation.is_none() && duplicate.remediation.is_some() {
-        primary.remediation.clone_from(&duplicate.remediation);
-    }
-}
-
 #[cfg(test)]
 mod tests {
    use super::*;
-    use compliance_core::models::dast::DastVulnType;
-    use compliance_core::models::finding::Severity;

    #[test]
    fn fingerprint_is_deterministic() {
@@ -257,159 +55,4 @@ mod tests {
        let b = compute_fingerprint(&["a", "bc"]);
        assert_ne!(a, b);
    }
-
-    fn make_dast(title: &str, endpoint: &str, vuln_type: DastVulnType) -> DastFinding {
-        let mut f = DastFinding::new(
-            "run1".into(),
-            "target1".into(),
-            vuln_type,
-            title.into(),
-            format!("Description for {title}"),
-            Severity::Medium,
-            endpoint.into(),
-            "GET".into(),
-        );
-        f.cwe = Some("CWE-319".into());
-        f
-    }
-
-    #[test]
-    fn canonicalize_strips_domain_suffix() {
-        let canon = canonicalize_dast_title("Missing HSTS header for comp-dev.meghsakha.com");
-        assert!(!canon.contains("meghsakha"), "domain should be stripped");
-        assert!(
-            canon.contains("strict-transport-security"),
-            "hsts should be resolved: {canon}"
-        );
-    }
-
-    #[test]
-    fn canonicalize_resolves_synonyms() {
-        let a = canonicalize_dast_title("Missing HSTS header");
-        let b = canonicalize_dast_title("Missing strict-transport-security header");
-        assert_eq!(a, b);
-    }
-
-    #[test]
-    fn exact_dedup_merges_identical_findings() {
-        let f1 = make_dast(
-            "Missing strict-transport-security header",
-            "https://example.com",
-            DastVulnType::SecurityHeaderMissing,
-        );
-        let f2 = make_dast(
-            "Missing strict-transport-security header",
-            "https://example.com",
-            DastVulnType::SecurityHeaderMissing,
-        );
-        let result = dedup_dast_findings(vec![f1, f2]);
-        assert_eq!(result.len(), 1, "exact duplicates should be merged");
-    }
-
-    #[test]
-    fn synonym_dedup_merges_hsts_variants() {
-        let f1 = make_dast(
-            "Missing strict-transport-security header",
-            "https://example.com",
-            DastVulnType::SecurityHeaderMissing,
-        );
-        let f2 = make_dast(
-            "Missing HSTS header for example.com",
-            "https://example.com",
-            DastVulnType::TlsMisconfiguration,
-        );
-        let result = dedup_dast_findings(vec![f1, f2]);
-        assert_eq!(
-            result.len(),
-            1,
-            "HSTS synonym variants should merge to 1 finding"
-        );
-    }
-
-    #[test]
-    fn different_headers_not_merged() {
-        let mut f1 = make_dast(
-            "Missing x-content-type-options header",
-            "https://example.com",
-            DastVulnType::SecurityHeaderMissing,
-        );
-        f1.cwe = Some("CWE-16".into());
-        let mut f2 = make_dast(
-            "Missing permissions-policy header",
-            "https://example.com",
-            DastVulnType::SecurityHeaderMissing,
-        );
-        f2.cwe = Some("CWE-16".into());
-        // These share CWE-16 but are different headers — phase 2 will merge them
-        // since they share the same CWE+endpoint. This is acceptable because they
-        // have the same root cause (missing security headers configuration).
-        let result = dedup_dast_findings(vec![f1, f2]);
-        // CWE-based dedup will merge these into 1
-        assert!(
-            result.len() <= 2,
-            "same CWE+endpoint findings may be merged"
-        );
-    }
-
-    #[test]
-    fn different_endpoints_not_merged() {
-        let f1 = make_dast(
-            "Missing strict-transport-security header",
-            "https://example.com",
-            DastVulnType::SecurityHeaderMissing,
-        );
-        let f2 = make_dast(
-            "Missing strict-transport-security header",
-            "https://other.com",
-            DastVulnType::SecurityHeaderMissing,
-        );
-        let result = dedup_dast_findings(vec![f1, f2]);
-        assert_eq!(result.len(), 2, "different endpoints should not merge");
-    }
-
-    #[test]
-    fn dedup_preserves_highest_severity() {
-        let f1 = make_dast(
-            "Missing strict-transport-security header",
-            "https://example.com",
-            DastVulnType::SecurityHeaderMissing,
-        );
-        let mut f2 = make_dast(
-            "Missing strict-transport-security header",
-            "https://example.com",
-            DastVulnType::SecurityHeaderMissing,
-        );
-        f2.severity = Severity::High;
-        let result = dedup_dast_findings(vec![f1, f2]);
-        assert_eq!(result.len(), 1);
-        assert_eq!(result[0].severity, Severity::High);
-    }
-
-    #[test]
-    fn dedup_merges_evidence() {
-        let mut f1 = make_dast(
-            "Missing strict-transport-security header",
-            "https://example.com",
-            DastVulnType::SecurityHeaderMissing,
-        );
-        f1.evidence
-            .push(compliance_core::models::dast::DastEvidence {
-                request_method: "GET".into(),
-                request_url: "https://example.com".into(),
-                request_headers: None,
-                request_body: None,
-                response_status: 200,
-                response_headers: None,
-                response_snippet: Some("pass 1".into()),
-                screenshot_path: None,
-                payload: None,
-                response_time_ms: None,
-            });
-        let mut f2 = f1.clone();
-        f2.evidence[0].response_snippet = Some("pass 2".into());
-
-        let result = dedup_dast_findings(vec![f1, f2]);
-        assert_eq!(result.len(), 1);
-        assert_eq!(result[0].evidence.len(), 2, "evidence should be merged");
-    }
 }
@@ -1,152 +0,0 @@
-//! Firmware SBOM via tramiton.
-//!
-//! Phase 2 (full, the default): drive a **reproducible build** with tramiton's
-//! `NixBackend` — `analyze` → `seal_and_build` → a sealed lock whose libraries
-//! are pinned and whose firmware artifact carries a content hash — then render
-//! the SBOM from the lock plus deep binary SCA of pre-compiled inputs. This is
-//! the complete bill of materials (toolchain + every fetched library + the
-//! firmware image), the same one `tramiton sbom` produces.
-//!
-//! Phase 1 fallback (analysis-only): when no nix backend is available or the
-//! build fails, fall back to the resolvable libraries + toolchain from the build
-//! plan alone (no build). A scan therefore always yields *something*, and a nix
-//! that can't run in the deployment never breaks a scan.
-
-use std::path::Path;
-
-use compliance_core::models::{SbomEntry, TargetType};
-use tramiton_repro::ReproBackend;
-use tramiton_sbom::ComponentKind;
-
-/// Whether firmware SBOM applies to this target family.
-pub fn is_firmware_target(target_type: TargetType) -> bool {
-    matches!(
-        target_type,
-        TargetType::FirmwareBareMetal | TargetType::FirmwareRtos | TargetType::EmbeddedLinuxYocto
-    )
-}
-
-/// Build SBOM entries for a firmware target from its source tree. Prefers a full
-/// reproducible build (sealed lock); falls back to analysis-only. Returns an
-/// empty vector when tramiton cannot even form a build plan.
-pub async fn firmware_sbom_entries(path: &Path, repo_id: &str) -> Vec<SbomEntry> {
-    let p = path.to_path_buf();
-    let repo = repo_id.to_string();
-    // The whole analyze → seal → build → render sequence is blocking (it shells
-    // out to nix), so keep it off the async runtime. Bound it: a firmware build
-    // that hangs must not wedge the scan (the orphaned task is abandoned).
-    let handle = tokio::task::spawn_blocking(move || build_sbom_blocking(&p, &repo));
-    match tokio::time::timeout(std::time::Duration::from_secs(900), handle).await {
-        Ok(Ok(entries)) => entries,
-        Ok(Err(e)) => {
-            tracing::warn!(repo_id, error = %e, "Firmware SBOM: task join error");
-            Vec::new()
-        }
-        Err(_) => {
-            tracing::warn!(repo_id, "Firmware SBOM: build exceeded 15m; skipping");
-            Vec::new()
-        }
-    }
-}
-
-fn build_sbom_blocking(path: &Path, repo_id: &str) -> Vec<SbomEntry> {
-    let repo = tramiton_core::Repo::new(path);
-    let plan = match tramiton_core::provider::analyze(&repo) {
-        Ok(Some(bp)) => bp,
-        Ok(None) => return Vec::new(),
-        Err(e) => {
-            tracing::warn!(repo_id, error = %e, "Firmware SBOM: tramiton analyze failed");
-            return Vec::new();
-        }
-    };
-
-    // Phase 2: reproducible build → sealed lock → complete SBOM.
-    if let Some(backend) = tramiton_repro::NixBackend::detect() {
-        match tramiton_repro::seal_and_build(&backend, &plan, path) {
-            Ok(lock) => {
-                let mut sbom = tramiton_sbom::Sbom::from_lock(&lock, repo_id);
-                // Deep binary SCA of any pre-compiled inputs in the tree.
-                sbom.components.extend(tramiton_sbom::binary::scan(path));
-                let entries = sbom_to_entries(&sbom, repo_id);
-                tracing::info!(
-                    repo_id,
-                    backend = backend.name(),
-                    count = entries.len(),
-                    "Firmware SBOM: sealed reproducible build"
-                );
-                return entries;
-            }
-            Err(e) => {
-                tracing::warn!(repo_id, error = %e, "Firmware SBOM: reproducible build failed; falling back to analysis-only")
-            }
-        }
-    } else {
-        tracing::info!(
-            repo_id,
-            "Firmware SBOM: no nix backend available; analysis-only SBOM"
-        );
-    }
-
-    // Phase 1 fallback: analysis-only (toolchain + resolvable libraries).
-    analysis_entries(&plan, repo_id)
-}
-
-/// Map a rendered [`tramiton_sbom::Sbom`] (primary firmware + components) into
-/// our [`SbomEntry`] rows. Source-file (`File`) components are dropped — they are
-/// build inputs, not a dependency inventory.
-fn sbom_to_entries(sbom: &tramiton_sbom::Sbom, repo_id: &str) -> Vec<SbomEntry> {
-    let mut entries = Vec::new();
-    if let Some(primary) = &sbom.primary {
-        entries.push(component_to_entry(primary, repo_id));
-    }
-    for c in &sbom.components {
-        if matches!(c.kind, ComponentKind::File) {
-            continue;
-        }
-        entries.push(component_to_entry(c, repo_id));
-    }
-    entries
-}
-
-fn component_to_entry(c: &tramiton_sbom::Component, repo_id: &str) -> SbomEntry {
-    let manager = match c.kind {
-        ComponentKind::Firmware => "firmware",
-        ComponentKind::Library => "library",
-        ComponentKind::Toolchain => "toolchain",
-        ComponentKind::File => "file",
-    };
-    let mut entry = SbomEntry::new(
-        repo_id.to_string(),
-        c.name.clone(),
-        c.version.clone().unwrap_or_default(),
-        manager.to_string(),
-    );
-    entry.purl = c.source.clone();
-    entry
-}
-
-/// Analysis-only components from the build plan: the cross-toolchain plus the
-/// resolvable fetched libraries, without a build.
-fn analysis_entries(bp: &tramiton_core::BuildPlan, repo_id: &str) -> Vec<SbomEntry> {
-    let mut entries = Vec::new();
-    if let Some(id) = bp.toolchain.id.clone() {
-        let version = bp.toolchain.version.clone().unwrap_or_default();
-        entries.push(SbomEntry::new(
-            repo_id.to_string(),
-            id,
-            version,
-            "toolchain".to_string(),
-        ));
-    }
-    for lib in tramiton_repro::lock::libraries_from_inputs(&bp.inputs) {
-        let mut entry = SbomEntry::new(
-            repo_id.to_string(),
-            lib.name,
-            lib.revision,
-            "library".to_string(),
-        );
-        entry.purl = lib.source;
-        entries.push(entry);
-    }
-    entries
-}
@@ -80,10 +80,7 @@ impl GitOps {

    #[tracing::instrument(skip_all, fields(repo_name = %repo_name))]
    pub fn clone_or_fetch(&self, git_url: &str, repo_name: &str) -> Result<PathBuf, AgentError> {
-        // Names can contain slashes or other path-hostile characters (a target
-        // named after a repo path, say); collapse to one safe directory segment
-        // so the clone path never nests or breaks.
-        let repo_path = self.base_path.join(sanitize_repo_dir(repo_name));
+        let repo_path = self.base_path.join(repo_name);

        if repo_path.exists() {
            tracing::info!("fetching updates for existing repo");
@@ -138,7 +135,7 @@ impl GitOps {
    /// Build credentials from agent config + per-repo overrides
    pub fn make_repo_credentials(
        config: &compliance_core::AgentConfig,
-        repo: &crate::pipeline::repo_view::RepoView,
+        repo: &compliance_core::models::TrackedRepository,
    ) -> RepoCredentials {
        RepoCredentials {
            ssh_key_path: Some(config.ssh_key_path.clone()),
@@ -256,46 +253,3 @@ pub struct DiffFile {
    pub path: String,
    pub hunks: String,
 }
-
-/// Collapse a repository name into a single filesystem-safe directory segment.
-/// Names may carry slashes or other path-hostile characters (a target named
-/// after a repo path, for instance); those would otherwise nest or break the
-/// clone path, so map anything outside `[A-Za-z0-9._-]` to `_`.
-fn sanitize_repo_dir(name: &str) -> String {
-    let mapped: String = name
-        .chars()
-        .map(|c| {
-            if c.is_ascii_alphanumeric() || c == '-' || c == '_' || c == '.' {
-                c
-            } else {
-                '_'
-            }
-        })
-        .collect();
-    let trimmed = mapped.trim_matches(|c| c == '.' || c == '_');
-    if trimmed.is_empty() {
-        "repo".to_string()
-    } else {
-        trimmed.to_string()
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use super::sanitize_repo_dir;
-
-    #[test]
-    fn sanitizes_path_hostile_names() {
-        assert_eq!(
-            sanitize_repo_dir("zephyr-example-app"),
-            "zephyr-example-app"
-        );
-        assert_eq!(
-            sanitize_repo_dir("ChristianRinn/bare_metal_stm32f411xe"),
-            "ChristianRinn_bare_metal_stm32f411xe"
-        );
-        assert_eq!(sanitize_repo_dir("../../etc/passwd"), "etc_passwd");
-        assert_eq!(sanitize_repo_dir("a b:c"), "a_b_c");
-        assert_eq!(sanitize_repo_dir("///"), "repo");
-    }
-}
@@ -19,33 +19,26 @@ impl Scanner for GitleaksScanner {

    #[tracing::instrument(skip_all)]
    async fn scan(&self, repo_path: &Path, repo_id: &str) -> Result<ScanOutput, CoreError> {
-        let output = tokio::time::timeout(
-            std::time::Duration::from_secs(300),
-            tokio::process::Command::new("gitleaks")
-                .args([
-                    "detect",
-                    "--source",
-                    ".",
-                    "--report-format",
-                    "json",
-                    "--report-path",
-                    "/dev/stdout",
-                    "--no-banner",
-                    "--exit-code",
-                    "0",
-                ])
-                .current_dir(repo_path)
-                .output(),
-        )
-        .await
-        .map_err(|_| CoreError::Scanner {
-            scanner: "gitleaks".to_string(),
-            source: "timed out after 5 minutes".into(),
-        })?
-        .map_err(|e| CoreError::Scanner {
-            scanner: "gitleaks".to_string(),
-            source: Box::new(e),
-        })?;
+        let output = tokio::process::Command::new("gitleaks")
+            .args([
+                "detect",
+                "--source",
+                ".",
+                "--report-format",
+                "json",
+                "--report-path",
+                "/dev/stdout",
+                "--no-banner",
+                "--exit-code",
+                "0",
+            ])
+            .current_dir(repo_path)
+            .output()
+            .await
+            .map_err(|e| CoreError::Scanner {
+                scanner: "gitleaks".to_string(),
+                source: Box::new(e),
+            })?;

        if output.stdout.is_empty() {
            return Ok(ScanOutput::default());
@@ -1,6 +1,5 @@
 use mongodb::bson::doc;

-use crate::pipeline::repo_view::RepoView;
 use compliance_core::models::*;

 use super::orchestrator::{extract_base_url, PipelineOrchestrator};
@@ -11,7 +10,7 @@ use crate::trackers;
 impl PipelineOrchestrator {
    /// Build an issue tracker client from a repository's tracker configuration.
    /// Returns `None` if the repo has no tracker configured.
-    pub(super) fn build_tracker(&self, repo: &RepoView) -> Option<TrackerDispatch> {
+    pub(super) fn build_tracker(&self, repo: &TrackedRepository) -> Option<TrackerDispatch> {
        let tracker_type = repo.tracker_type.as_ref()?;
        // Per-repo token takes precedence, fall back to global config
        match tracker_type {
@@ -82,7 +81,7 @@ impl PipelineOrchestrator {
    #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
    pub(super) async fn create_tracker_issues(
        &self,
-        repo: &RepoView,
+        repo: &TrackedRepository,
        repo_id: &str,
        new_findings: &[Finding],
    ) -> Result<(), AgentError> {
@@ -1,7 +1,6 @@
 pub mod code_review;
 pub mod cve;
 pub mod dedup;
-pub mod firmware_sbom;
 pub mod git;
 pub mod gitleaks;
 mod graph_build;
@@ -9,10 +8,7 @@ mod issue_creation;
 pub mod lint;
 pub mod orchestrator;
 pub mod patterns;
-pub mod plan;
-pub mod plc;
 mod pr_review;
-pub mod repo_view;
 pub mod sbom;
 pub mod semgrep;
 mod tracker_dispatch;
@@ -33,7 +33,6 @@ struct PatternRule {
    file_extensions: Vec<String>,
 }

-#[allow(clippy::new_without_default)]
 impl GdprPatternScanner {
    pub fn new() -> Self {
        let patterns = vec![
@@ -99,7 +98,6 @@ impl Scanner for GdprPatternScanner {
    }
 }

-#[allow(clippy::new_without_default)]
 impl OAuthPatternScanner {
    pub fn new() -> Self {
        let patterns = vec![
@@ -215,7 +213,7 @@ fn scan_with_patterns(
                        repo_id.to_string(),
                        fingerprint,
                        scanner_name.to_string(),
-                        scan_type,
+                        scan_type.clone(),
                        pattern.title.clone(),
                        pattern.description.clone(),
                        pattern.severity.clone(),
@@ -1,214 +0,0 @@
-//! The scan plan.
-//!
-//! [`build_scan_plan`] turns an [`OnboardedTarget`] into the concrete ordered
-//! list of scans to run, each bound to the artifact it consumes. It intersects
-//! the scan-applicability matrix ([`applicable_scans`]) with the target's
-//! `scan_config` overrides: a scan runs when its required artifact is present and
-//! it is either on by default or explicitly enabled, and is not explicitly
-//! disabled. This is the decision engine the unified pipeline (`run_target`)
-//! executes.
-
-use compliance_core::models::{Artifact, ArtifactKind, OnboardedTarget, ScanPhase, ScanType};
-use compliance_core::scan_matrix::applicable_scans;
-
-/// One scan to run, bound to the artifact it operates on.
-#[derive(Debug, Clone, PartialEq)]
-pub struct ScanStep {
-    /// The scan to run.
-    pub scan_type: ScanType,
-    /// The pipeline phase to report while it runs.
-    pub phase: ScanPhase,
-    /// The id of the artifact this scan consumes ([`Artifact::id`]).
-    pub artifact_id: String,
-}
-
-/// The ordered set of scans to run for a target.
-#[derive(Debug, Clone, Default, PartialEq)]
-pub struct ScanPlan {
-    /// The scans, in matrix order.
-    pub steps: Vec<ScanStep>,
-}
-
-impl ScanPlan {
-    /// Whether the plan contains a step for the given scan type.
-    pub fn has(&self, scan: ScanType) -> bool {
-        self.steps.iter().any(|s| s.scan_type == scan)
-    }
-
-    /// Whether the plan is empty (nothing to run).
-    pub fn is_empty(&self) -> bool {
-        self.steps.is_empty()
-    }
-}
-
-/// Build the scan plan for a target: matrix defaults ∩ `scan_config`, each scan
-/// bound to the artifact it consumes. Scans whose required artifact is absent, or
-/// that are disabled, or off-by-default and not explicitly enabled, are dropped.
-pub fn build_scan_plan(target: &OnboardedTarget) -> ScanPlan {
-    let enabled = &target.scan_config.enabled_scans;
-    let disabled = &target.scan_config.disabled_scans;
-
-    let mut steps = Vec::new();
-    for option in applicable_scans(target) {
-        // Required artifact missing → not runnable.
-        if option.blocked_reason.is_some() {
-            continue;
-        }
-        // Explicit opt-out wins.
-        if disabled.contains(&option.scan) {
-            continue;
-        }
-        // Run if on by default, or explicitly enabled.
-        if !option.default_on && !enabled.contains(&option.scan) {
-            continue;
-        }
-        let Some(artifact) = resolve_artifact(target, option.required_artifact) else {
-            continue;
-        };
-        steps.push(ScanStep {
-            scan_type: option.scan,
-            phase: phase_for(option.scan),
-            artifact_id: artifact.id.clone(),
-        });
-    }
-    ScanPlan { steps }
-}
-
-/// Resolve the artifact a scan consumes. A "code" requirement (represented by
-/// `GitRepo`) is satisfied by a git repo *or* a source archive. The PLC
-/// control-logic requirement (represented by `PlcProject`) prefers an uploaded
-/// PLC project but also accepts a code artifact — a git repo / source archive
-/// holding PLCopen XML / ST exports.
-fn resolve_artifact(target: &OnboardedTarget, required: Option<ArtifactKind>) -> Option<&Artifact> {
-    match required {
-        Some(ArtifactKind::GitRepo) => target.code_artifact(),
-        Some(ArtifactKind::PlcProject) => target
-            .first_of(ArtifactKind::PlcProject)
-            .or_else(|| target.code_artifact()),
-        Some(kind) => target.first_of(kind),
-        None => target.code_artifact().or_else(|| target.artifacts.first()),
-    }
-}
-
-/// The pipeline phase reported while a given scan runs.
-fn phase_for(scan: ScanType) -> ScanPhase {
-    match scan {
-        ScanType::Sast => ScanPhase::Sast,
-        ScanType::Sbom => ScanPhase::SbomGeneration,
-        ScanType::Cve => ScanPhase::CveScanning,
-        ScanType::Gdpr | ScanType::OAuth => ScanPhase::PatternScanning,
-        ScanType::SecretDetection => ScanPhase::SecretDetection,
-        ScanType::Lint => ScanPhase::LintScanning,
-        ScanType::CodeReview => ScanPhase::CodeReview,
-        ScanType::Graph => ScanPhase::GraphBuilding,
-        ScanType::Dast => ScanPhase::DastScanning,
-        ScanType::FirmwareStatic => ScanPhase::FirmwareStatic,
-        ScanType::PlcControlLogic => ScanPhase::PlcAnalysis,
-        ScanType::MobileStatic => ScanPhase::MobileStatic,
-        ScanType::ContainerScan => ScanPhase::ContainerScan,
-        ScanType::IcsProbe => ScanPhase::IcsProbe,
-    }
-}
-
-#[cfg(test)]
-#[allow(clippy::expect_used, clippy::unwrap_used)]
-mod tests {
-    use super::*;
-    use compliance_core::models::{PlcFormat, TargetType};
-
-    fn target(target_type: TargetType, artifacts: Vec<Artifact>) -> OnboardedTarget {
-        let mut t = OnboardedTarget::new("t".to_string(), target_type);
-        t.artifacts = artifacts;
-        t
-    }
-
-    fn step_for<'a>(plan: &'a ScanPlan, scan: ScanType) -> Option<&'a ScanStep> {
-        plan.steps.iter().find(|s| s.scan_type == scan)
-    }
-
-    #[test]
-    fn webapp_with_code_and_url_runs_sast_and_dast_bound_to_the_right_artifacts() {
-        let code = Artifact::git_repo("https://git/x", "main");
-        let url = Artifact::live_url("https://x");
-        let (code_id, url_id) = (code.id.clone(), url.id.clone());
-        let t = target(TargetType::WebApp, vec![code, url]);
-
-        let plan = build_scan_plan(&t);
-        let sast = step_for(&plan, ScanType::Sast).expect("sast planned");
-        assert_eq!(sast.artifact_id, code_id);
-        assert_eq!(sast.phase, ScanPhase::Sast);
-        let dast = step_for(&plan, ScanType::Dast).expect("dast planned");
-        assert_eq!(dast.artifact_id, url_id);
-    }
-
-    #[test]
-    fn webapp_without_url_omits_dast() {
-        let t = target(TargetType::WebApp, vec![Artifact::git_repo("u", "main")]);
-        let plan = build_scan_plan(&t);
-        assert!(plan.has(ScanType::Sast));
-        assert!(!plan.has(ScanType::Dast));
-    }
-
-    #[test]
-    fn code_scan_binds_to_source_archive_when_no_git_repo() {
-        let arc = Artifact::source_archive("src.zip");
-        let arc_id = arc.id.clone();
-        let t = target(TargetType::BackendService, vec![arc]);
-        let plan = build_scan_plan(&t);
-        let sast = step_for(&plan, ScanType::Sast).expect("sast planned");
-        assert_eq!(sast.artifact_id, arc_id);
-    }
-
-    #[test]
-    fn firmware_sbom_and_cve_bind_to_the_firmware_image() {
-        let fw = Artifact::firmware_image("fw.bin");
-        let fw_id = fw.id.clone();
-        let t = target(TargetType::FirmwareBareMetal, vec![fw]);
-        let plan = build_scan_plan(&t);
-        let sbom = step_for(&plan, ScanType::Sbom).expect("sbom planned");
-        assert_eq!(sbom.artifact_id, fw_id);
-        assert!(step_for(&plan, ScanType::FirmwareStatic).is_some());
-        assert!(!plan.has(ScanType::Dast));
-    }
-
-    #[test]
-    fn plc_plans_only_control_logic() {
-        let plc = Artifact::plc_project("p.xml", PlcFormat::PlcopenXml);
-        let t = target(TargetType::PlcSps, vec![plc]);
-        let plan = build_scan_plan(&t);
-        assert_eq!(plan.steps.len(), 1);
-        assert_eq!(plan.steps[0].scan_type, ScanType::PlcControlLogic);
-        assert_eq!(plan.steps[0].phase, ScanPhase::PlcAnalysis);
-    }
-
-    #[test]
-    fn plc_control_logic_binds_to_a_git_repo() {
-        // A CODESYS project in git (PLCopen XML / ST exports) with no uploaded
-        // PlcProject: control-logic still plans, bound to the git artifact.
-        let git = Artifact::git_repo("https://git/plc", "main");
-        let git_id = git.id.clone();
-        let t = target(TargetType::PlcSps, vec![git]);
-        let plan = build_scan_plan(&t);
-        let step = step_for(&plan, ScanType::PlcControlLogic).expect("control-logic planned");
-        assert_eq!(step.artifact_id, git_id, "PLC scan binds to the git repo");
-    }
-
-    #[test]
-    fn disabled_scan_is_dropped_and_off_by_default_can_be_enabled() {
-        let mut t = target(TargetType::WebApp, vec![Artifact::git_repo("u", "main")]);
-        t.scan_config.disabled_scans = vec![ScanType::Lint];
-        // CodeReview is off by default for web; enable it explicitly.
-        t.scan_config.enabled_scans = vec![ScanType::CodeReview];
-        let plan = build_scan_plan(&t);
-        assert!(!plan.has(ScanType::Lint));
-        assert!(plan.has(ScanType::CodeReview));
-        assert!(plan.has(ScanType::Sast));
-    }
-
-    #[test]
-    fn no_code_artifact_yields_empty_plan_for_web() {
-        let t = target(TargetType::WebApp, vec![]);
-        let plan = build_scan_plan(&t);
-        assert!(plan.is_empty());
-    }
-}
@@ -1,226 +0,0 @@
-//! Abstract syntax tree for IEC 61131-3 Structured Text (ST).
-//!
-//! This is the security-relevant subset: POUs with their variable declarations
-//! and statement bodies, enough to run semantic control-logic rules over. It is
-//! deliberately not a full language model — declarations we don't reason about
-//! (e.g. exotic type definitions) are parsed loosely and kept as raw text.
-
-/// A Program Organization Unit: a PROGRAM, FUNCTION, or FUNCTION_BLOCK.
-#[derive(Debug, Clone)]
-pub struct Pou {
-    pub name: String,
-    pub kind: PouKind,
-    /// The declared variables, across all VAR_* sections.
-    pub vars: Vec<VarDecl>,
-    /// The statement body.
-    pub body: Vec<Stmt>,
-    /// 1-based line where the POU header appears (in the source that was parsed).
-    pub line: u32,
-}
-
-#[derive(Debug, Clone, Copy, PartialEq, Eq)]
-pub enum PouKind {
-    Program,
-    Function,
-    FunctionBlock,
-}
-
-impl PouKind {
-    pub fn label(self) -> &'static str {
-        match self {
-            PouKind::Program => "PROGRAM",
-            PouKind::Function => "FUNCTION",
-            PouKind::FunctionBlock => "FUNCTION_BLOCK",
-        }
-    }
-}
-
-/// A single declared variable.
-#[derive(Debug, Clone)]
-pub struct VarDecl {
-    pub name: String,
-    pub section: VarSection,
-    /// The declared type as written (e.g. `BOOL`, `INT`, `ARRAY[0..9] OF INT`).
-    pub type_name: String,
-    /// Whether the type is an ARRAY, and its declared bounds `(lo, hi)` when
-    /// they are literal integers — used by the array-bounds rule.
-    pub array_bounds: Option<(i64, i64)>,
-    /// The initializer expression, if any (`:= <expr>`).
-    pub init: Option<Expr>,
-    pub line: u32,
-}
-
-#[derive(Debug, Clone, Copy, PartialEq, Eq)]
-pub enum VarSection {
-    Var,
-    Input,
-    Output,
-    InOut,
-    Global,
-    Temp,
-    External,
-}
-
-/// A statement.
-#[derive(Debug, Clone)]
-pub enum Stmt {
-    Assign {
-        target: Expr,
-        value: Expr,
-        line: u32,
-    },
-    If {
-        /// (condition, body) for IF and each ELSIF, in order.
-        branches: Vec<(Expr, Vec<Stmt>)>,
-        else_body: Option<Vec<Stmt>>,
-        line: u32,
-    },
-    Case {
-        selector: Expr,
-        /// (label expressions, body) per CASE arm.
-        arms: Vec<(Vec<Expr>, Vec<Stmt>)>,
-        else_body: Option<Vec<Stmt>>,
-        line: u32,
-    },
-    For {
-        var: String,
-        from: Expr,
-        to: Expr,
-        by: Option<Expr>,
-        body: Vec<Stmt>,
-        line: u32,
-    },
-    While {
-        cond: Expr,
-        body: Vec<Stmt>,
-        line: u32,
-    },
-    Repeat {
-        body: Vec<Stmt>,
-        until: Expr,
-        line: u32,
-    },
-    /// A bare call statement, e.g. `TON1(IN := x, PT := T#5s);`.
-    Call {
-        callee: String,
-        args: Vec<CallArg>,
-        line: u32,
-    },
-    Return {
-        line: u32,
-    },
-    Exit {
-        line: u32,
-    },
-    /// `JMP label;` — an unstructured jump.
-    Jump {
-        label: String,
-        line: u32,
-    },
-    /// `label:` — a jump target.
-    Label {
-        name: String,
-        line: u32,
-    },
-}
-
-/// One argument in a call: positional (`name: None`) or named (`X := expr`).
-#[derive(Debug, Clone)]
-pub struct CallArg {
-    pub name: Option<String>,
-    pub value: Expr,
-}
-
-/// An expression.
-#[derive(Debug, Clone)]
-pub enum Expr {
-    Int(i64, u32),
-    Real(f64, u32),
-    Bool(bool, u32),
-    /// A string literal, with the unquoted contents.
-    Str(String, u32),
-    /// A duration / date / time literal, kept as raw text (`T#5s`, `DT#...`).
-    Time(String, u32),
-    Ident(String, u32),
-    /// `base[index]`.
-    Index {
-        base: Box<Expr>,
-        index: Box<Expr>,
-        line: u32,
-    },
-    /// `base.field`.
-    Member {
-        base: Box<Expr>,
-        field: String,
-        line: u32,
-    },
-    Unary {
-        op: UnOp,
-        expr: Box<Expr>,
-        line: u32,
-    },
-    Binary {
-        op: BinOp,
-        lhs: Box<Expr>,
-        rhs: Box<Expr>,
-        line: u32,
-    },
-    /// A function call used as an expression, e.g. `LIMIT(a, b, c)`.
-    Call {
-        callee: String,
-        args: Vec<CallArg>,
-        line: u32,
-    },
-}
-
-impl Expr {
-    /// The 1-based source line this expression starts on.
-    pub fn line(&self) -> u32 {
-        match self {
-            Expr::Int(_, l)
-            | Expr::Real(_, l)
-            | Expr::Bool(_, l)
-            | Expr::Str(_, l)
-            | Expr::Time(_, l)
-            | Expr::Ident(_, l)
-            | Expr::Index { line: l, .. }
-            | Expr::Member { line: l, .. }
-            | Expr::Unary { line: l, .. }
-            | Expr::Binary { line: l, .. }
-            | Expr::Call { line: l, .. } => *l,
-        }
-    }
-
-    /// If this expression is a plain identifier, its name.
-    pub fn as_ident(&self) -> Option<&str> {
-        match self {
-            Expr::Ident(name, _) => Some(name.as_str()),
-            _ => None,
-        }
-    }
-}
-
-#[derive(Debug, Clone, Copy, PartialEq, Eq)]
-pub enum UnOp {
-    Not,
-    Neg,
-}
-
-#[derive(Debug, Clone, Copy, PartialEq, Eq)]
-pub enum BinOp {
-    Add,
-    Sub,
-    Mul,
-    Div,
-    Mod,
-    Pow,
-    Eq,
-    Ne,
-    Lt,
-    Le,
-    Gt,
-    Ge,
-    And,
-    Or,
-    Xor,
-}
@@ -1,372 +0,0 @@
-//! Lexer for IEC 61131-3 Structured Text.
-//!
-//! Tokenizes ST source into a flat token stream with 1-based line numbers.
-//! Keywords are case-insensitive. Handles `(* *)` and `//` comments, `'..'` and
-//! `".."` strings (with `''`/`""` escapes), based integers (`16#FF`, `2#1010`),
-//! and duration/date literals (`T#5s`, `DT#...`) kept as raw text.
-
-/// A lexed token with its source line.
-#[derive(Debug, Clone)]
-pub struct Token {
-    pub kind: Tok,
-    pub line: u32,
-}
-
-#[derive(Debug, Clone, PartialEq)]
-pub enum Tok {
-    Int(i64),
-    Real(f64),
-    Str(String),
-    Time(String),
-    Bool(bool),
-    Ident(String),
-    Kw(Keyword),
-    Assign, // :=
-    Plus,   // +
-    Minus,  // -
-    Star,   // *
-    Slash,  // /
-    Power,  // **
-    LParen, // (
-    RParen, // )
-    LBrack, // [
-    RBrack, // ]
-    Dot,    // .
-    DotDot, // ..
-    Comma,  // ,
-    Semi,   // ;
-    Colon,  // :
-    Lt,     // <
-    Le,     // <=
-    Gt,     // >
-    Ge,     // >=
-    Eq,     // =
-    Ne,     // <>
-    Amp,    // &
-    Eof,
-}
-
-#[derive(Debug, Clone, Copy, PartialEq, Eq)]
-pub enum Keyword {
-    Program,
-    EndProgram,
-    Function,
-    EndFunction,
-    FunctionBlock,
-    EndFunctionBlock,
-    Var,
-    VarInput,
-    VarOutput,
-    VarInOut,
-    VarGlobal,
-    VarTemp,
-    VarExternal,
-    Constant,
-    EndVar,
-    Array,
-    Of,
-    If,
-    Then,
-    Elsif,
-    Else,
-    EndIf,
-    Case,
-    EndCase,
-    For,
-    To,
-    By,
-    Do,
-    EndFor,
-    While,
-    EndWhile,
-    Repeat,
-    Until,
-    EndRepeat,
-    Return,
-    Exit,
-    Jmp,
-    Not,
-    And,
-    Or,
-    Xor,
-    Mod,
-    Type,
-    EndType,
-    Struct,
-    EndStruct,
-}
-
-fn keyword_from(word: &str) -> Option<Keyword> {
-    use Keyword::*;
-    Some(match word.to_ascii_uppercase().as_str() {
-        "PROGRAM" => Program,
-        "END_PROGRAM" => EndProgram,
-        "FUNCTION" => Function,
-        "END_FUNCTION" => EndFunction,
-        "FUNCTION_BLOCK" => FunctionBlock,
-        "END_FUNCTION_BLOCK" => EndFunctionBlock,
-        "VAR" => Var,
-        "VAR_INPUT" => VarInput,
-        "VAR_OUTPUT" => VarOutput,
-        "VAR_IN_OUT" => VarInOut,
-        "VAR_GLOBAL" => VarGlobal,
-        "VAR_TEMP" => VarTemp,
-        "VAR_EXTERNAL" => VarExternal,
-        "CONSTANT" => Constant,
-        "END_VAR" => EndVar,
-        "ARRAY" => Array,
-        "OF" => Of,
-        "IF" => If,
-        "THEN" => Then,
-        "ELSIF" => Elsif,
-        "ELSE" => Else,
-        "END_IF" => EndIf,
-        "CASE" => Case,
-        "END_CASE" => EndCase,
-        "FOR" => For,
-        "TO" => To,
-        "BY" => By,
-        "DO" => Do,
-        "END_FOR" => EndFor,
-        "WHILE" => While,
-        "END_WHILE" => EndWhile,
-        "REPEAT" => Repeat,
-        "UNTIL" => Until,
-        "END_REPEAT" => EndRepeat,
-        "RETURN" => Return,
-        "EXIT" => Exit,
-        "JMP" => Jmp,
-        "NOT" => Not,
-        "AND" => And,
-        "OR" => Or,
-        "XOR" => Xor,
-        "MOD" => Mod,
-        "TYPE" => Type,
-        "END_TYPE" => EndType,
-        "STRUCT" => Struct,
-        "END_STRUCT" => EndStruct,
-        _ => return None,
-    })
-}
-
-/// Tokenize `src`. Unknown characters are skipped (best-effort — a scanner must
-/// not die on odd input).
-pub fn lex(src: &str) -> Vec<Token> {
-    let chars: Vec<char> = src.chars().collect();
-    let mut i = 0usize;
-    let mut line = 1u32;
-    let mut out = Vec::new();
-
-    let bump_line = |c: char, line: &mut u32| {
-        if c == '\n' {
-            *line += 1;
-        }
-    };
-
-    while i < chars.len() {
-        let c = chars[i];
-
-        // Whitespace.
-        if c.is_whitespace() {
-            bump_line(c, &mut line);
-            i += 1;
-            continue;
-        }
-
-        // Line comment: //
-        if c == '/' && i + 1 < chars.len() && chars[i + 1] == '/' {
-            while i < chars.len() && chars[i] != '\n' {
-                i += 1;
-            }
-            continue;
-        }
-
-        // Block comment: (* ... *)
-        if c == '(' && i + 1 < chars.len() && chars[i + 1] == '*' {
-            i += 2;
-            while i + 1 < chars.len() && !(chars[i] == '*' && chars[i + 1] == ')') {
-                bump_line(chars[i], &mut line);
-                i += 1;
-            }
-            i = (i + 2).min(chars.len());
-            continue;
-        }
-
-        let tok_line = line;
-
-        // String literal: '...' or "..."
-        if c == '\'' || c == '"' {
-            let quote = c;
-            i += 1;
-            let mut s = String::new();
-            while i < chars.len() {
-                let ch = chars[i];
-                if ch == quote {
-                    // Doubled quote is an escaped quote.
-                    if i + 1 < chars.len() && chars[i + 1] == quote {
-                        s.push(quote);
-                        i += 2;
-                        continue;
-                    }
-                    i += 1;
-                    break;
-                }
-                bump_line(ch, &mut line);
-                s.push(ch);
-                i += 1;
-            }
-            out.push(Token {
-                kind: Tok::Str(s),
-                line: tok_line,
-            });
-            continue;
-        }
-
-        // Identifier / keyword / time literal / boolean.
-        if c.is_ascii_alphabetic() || c == '_' {
-            let start = i;
-            while i < chars.len() && (chars[i].is_ascii_alphanumeric() || chars[i] == '_') {
-                i += 1;
-            }
-            let word: String = chars[start..i].iter().collect();
-
-            // Duration/date/time literal prefix: T#, TIME#, DT#, D#, TOD#, LT# ...
-            if i < chars.len() && chars[i] == '#' {
-                let up = word.to_ascii_uppercase();
-                if matches!(
-                    up.as_str(),
-                    "T" | "TIME" | "DT" | "D" | "TOD" | "LT" | "DATE"
-                ) {
-                    let lit_start = start;
-                    i += 1; // consume '#'
-                    while i < chars.len()
-                        && (chars[i].is_ascii_alphanumeric()
-                            || chars[i] == '.'
-                            || chars[i] == '_'
-                            || chars[i] == ':')
-                    {
-                        i += 1;
-                    }
-                    let lit: String = chars[lit_start..i].iter().collect();
-                    out.push(Token {
-                        kind: Tok::Time(lit),
-                        line: tok_line,
-                    });
-                    continue;
-                }
-            }
-
-            let kind = match word.to_ascii_uppercase().as_str() {
-                "TRUE" => Tok::Bool(true),
-                "FALSE" => Tok::Bool(false),
-                _ => match keyword_from(&word) {
-                    Some(kw) => Tok::Kw(kw),
-                    None => Tok::Ident(word),
-                },
-            };
-            out.push(Token {
-                kind,
-                line: tok_line,
-            });
-            continue;
-        }
-
-        // Number: decimal, real, or based (16#..., 2#...).
-        if c.is_ascii_digit() {
-            let start = i;
-            while i < chars.len() && (chars[i].is_ascii_digit() || chars[i] == '_') {
-                i += 1;
-            }
-            // Based literal: <base>#<digits>
-            if i < chars.len() && chars[i] == '#' {
-                let base_str: String = chars[start..i].iter().filter(|c| **c != '_').collect();
-                i += 1;
-                let dstart = i;
-                while i < chars.len() && (chars[i].is_ascii_alphanumeric() || chars[i] == '_') {
-                    i += 1;
-                }
-                let digits: String = chars[dstart..i].iter().filter(|c| **c != '_').collect();
-                let radix = base_str.parse::<u32>().unwrap_or(10);
-                let val = i64::from_str_radix(&digits, radix.clamp(2, 36)).unwrap_or(0);
-                out.push(Token {
-                    kind: Tok::Int(val),
-                    line: tok_line,
-                });
-                continue;
-            }
-            // Real: has a '.' (not '..') or exponent.
-            let is_real =
-                i < chars.len() && chars[i] == '.' && !(i + 1 < chars.len() && chars[i + 1] == '.');
-            if is_real {
-                i += 1;
-                while i < chars.len() && (chars[i].is_ascii_digit() || chars[i] == '_') {
-                    i += 1;
-                }
-                let raw: String = chars[start..i].iter().filter(|c| **c != '_').collect();
-                out.push(Token {
-                    kind: Tok::Real(raw.parse().unwrap_or(0.0)),
-                    line: tok_line,
-                });
-                continue;
-            }
-            let raw: String = chars[start..i].iter().filter(|c| **c != '_').collect();
-            out.push(Token {
-                kind: Tok::Int(raw.parse().unwrap_or(0)),
-                line: tok_line,
-            });
-            continue;
-        }
-
-        // Operators / punctuation (longest match first).
-        let two: String = chars[i..(i + 2).min(chars.len())].iter().collect();
-        let kind = match two.as_str() {
-            ":=" => Some(Tok::Assign),
-            "<=" => Some(Tok::Le),
-            ">=" => Some(Tok::Ge),
-            "<>" => Some(Tok::Ne),
-            ".." => Some(Tok::DotDot),
-            "**" => Some(Tok::Power),
-            _ => None,
-        };
-        if let Some(k) = kind {
-            out.push(Token {
-                kind: k,
-                line: tok_line,
-            });
-            i += 2;
-            continue;
-        }
-        let one = match c {
-            '+' => Some(Tok::Plus),
-            '-' => Some(Tok::Minus),
-            '*' => Some(Tok::Star),
-            '/' => Some(Tok::Slash),
-            '(' => Some(Tok::LParen),
-            ')' => Some(Tok::RParen),
-            '[' => Some(Tok::LBrack),
-            ']' => Some(Tok::RBrack),
-            '.' => Some(Tok::Dot),
-            ',' => Some(Tok::Comma),
-            ';' => Some(Tok::Semi),
-            ':' => Some(Tok::Colon),
-            '<' => Some(Tok::Lt),
-            '>' => Some(Tok::Gt),
-            '=' => Some(Tok::Eq),
-            '&' => Some(Tok::Amp),
-            _ => None,
-        };
-        if let Some(k) = one {
-            out.push(Token {
-                kind: k,
-                line: tok_line,
-            });
-        }
-        i += 1;
-    }
-
-    out.push(Token {
-        kind: Tok::Eof,
-        line,
-    });
-    out
-}
@@ -1,234 +0,0 @@
-//! PLC control-logic security scanner for IEC 61131-3 targets.
-//!
-//! Parses Structured Text (raw `.st`/`.scl`/`.exp` files and PLCopen-XML
-//! projects) into an AST and runs semantic control-logic security rules over it.
-//! Implements [`ScanType::PlcControlLogic`].
-
-pub mod ast;
-pub mod lexer;
-pub mod parser;
-pub mod plcopen;
-pub mod rules;
-pub mod sbom;
-
-use std::path::Path;
-
-use compliance_core::error::CoreError;
-use compliance_core::models::{Finding, ScanType};
-use compliance_core::traits::{ScanOutput, Scanner};
-
-use crate::pipeline::dedup;
-
-/// Scanner for `ScanType::PlcControlLogic`.
-pub struct PlcControlLogicScanner;
-
-impl Scanner for PlcControlLogicScanner {
-    fn name(&self) -> &str {
-        "plc-control-logic"
-    }
-
-    fn scan_type(&self) -> ScanType {
-        ScanType::PlcControlLogic
-    }
-
-    #[tracing::instrument(skip_all)]
-    async fn scan(&self, repo_path: &Path, repo_id: &str) -> Result<ScanOutput, CoreError> {
-        let findings = analyze_tree(repo_path, repo_id);
-        Ok(ScanOutput {
-            findings,
-            sbom_entries: Vec::new(),
-        })
-    }
-}
-
-/// Walk a PLC project tree and produce findings.
-pub(crate) fn analyze_tree(root: &Path, repo_id: &str) -> Vec<Finding> {
-    let mut findings = Vec::new();
-    for entry in walkdir::WalkDir::new(root)
-        .into_iter()
-        .filter_map(|e| e.ok())
-    {
-        if !entry.file_type().is_file() {
-            continue;
-        }
-        let path = entry.path();
-        let ext = path
-            .extension()
-            .and_then(|e| e.to_str())
-            .unwrap_or("")
-            .to_ascii_lowercase();
-        let is_st = matches!(ext.as_str(), "st" | "iecst" | "scl" | "exp" | "il");
-        let is_xml = matches!(ext.as_str(), "xml" | "plcopen" | "project");
-        if !is_st && !is_xml {
-            continue;
-        }
-        let Ok(content) = std::fs::read_to_string(path) else {
-            continue;
-        };
-        let pous = if is_xml {
-            plcopen::parse_plcopen(&content)
-        } else {
-            parser::parse(&content)
-        };
-        if pous.is_empty() {
-            continue;
-        }
-        let rel = path
-            .strip_prefix(root)
-            .unwrap_or(path)
-            .to_string_lossy()
-            .to_string();
-        for pou in &pous {
-            for hit in rules::analyze(pou) {
-                let line_s = hit.line.to_string();
-                let fingerprint =
-                    dedup::compute_fingerprint(&[repo_id, &rel, hit.rule_id, &pou.name, &line_s]);
-                let mut f = Finding::new(
-                    repo_id.to_string(),
-                    fingerprint,
-                    "plc-control-logic".to_string(),
-                    ScanType::PlcControlLogic,
-                    hit.title,
-                    hit.description,
-                    hit.severity,
-                );
-                f.file_path = Some(rel.clone());
-                f.line_number = Some(hit.line);
-                f.rule_id = Some(hit.rule_id.to_string());
-                f.cwe = hit.cwe.map(String::from);
-                f.remediation = Some(hit.remediation.to_string());
-                findings.push(f);
-            }
-        }
-    }
-    findings
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use std::collections::HashSet;
-    use std::path::PathBuf;
-
-    fn demo_dir() -> PathBuf {
-        PathBuf::from(env!("CARGO_MANIFEST_DIR"))
-            .parent()
-            .expect("workspace root")
-            .join("examples/plc-demo")
-    }
-
-    #[test]
-    fn scans_demo_project_end_to_end() {
-        let findings = analyze_tree(&demo_dir(), "demo-target");
-        assert!(!findings.is_empty(), "demo project should produce findings");
-
-        let rules: HashSet<&str> = findings
-            .iter()
-            .filter_map(|f| f.rule_id.as_deref())
-            .collect();
-        for r in [
-            "plc-hardcoded-credential",
-            "plc-default-password",
-            "plc-safety-bypass",
-            "plc-array-unchecked-index",
-            "plc-insecure-comm",
-            "plc-insecure-protocol-port",
-            "plc-unstructured-jump",
-            "plc-division-by-zero",
-        ] {
-            assert!(rules.contains(r), "expected rule {r}; got {rules:?}");
-        }
-
-        // Every finding is well-formed for storage.
-        for f in &findings {
-            assert_eq!(f.repo_id, "demo-target");
-            assert!(f.file_path.is_some(), "finding needs a file");
-            assert!(f.line_number.is_some(), "finding needs a line");
-        }
-
-        // The guarded division (IF ScaleFactor <> 0.0) must not be double-counted:
-        // exactly one division-by-zero (the unguarded MeasuredFlow divide).
-        let div0 = findings
-            .iter()
-            .filter(|f| f.rule_id.as_deref() == Some("plc-division-by-zero"))
-            .count();
-        assert_eq!(div0, 1, "only the unguarded division should be flagged");
-    }
-
-    /// The realistic OpenPLC-style traffic-light sample is mostly sound control
-    /// logic: the scanner must surface its few genuine defects and stay quiet on
-    /// the timed state machine and the guarded duty-cycle division.
-    #[test]
-    fn realistic_sample_flags_only_real_issues() {
-        let all = analyze_tree(&demo_dir(), "demo-target");
-        let tl: Vec<_> = all
-            .iter()
-            .filter(|f| {
-                f.file_path
-                    .as_deref()
-                    .is_some_and(|p| p.ends_with("traffic_light.st"))
-            })
-            .collect();
-        assert!(!tl.is_empty(), "traffic_light.st should produce findings");
-
-        let rules: HashSet<&str> = tl.iter().filter_map(|f| f.rule_id.as_deref()).collect();
-        // The three planted defects: hardcoded SCADA password, cleartext Modbus
-        // master (no auth), and a maintenance mode that drops the PedPermit.
-        for r in [
-            "plc-hardcoded-credential",
-            "plc-insecure-comm",
-            "plc-safety-bypass",
-        ] {
-            assert!(rules.contains(r), "expected rule {r}; got {rules:?}");
-        }
-        // Modbus/TCP on 502 is also an insecure-protocol port.
-        assert!(rules.contains("plc-insecure-protocol-port"));
-
-        // Low false positives: the guarded `IF LampCount <> 0` division and the
-        // JMP-free state machine must not trip anything.
-        assert_eq!(
-            tl.iter()
-                .filter(|f| f.rule_id.as_deref() == Some("plc-division-by-zero"))
-                .count(),
-            0,
-            "the guarded duty-cycle division must not be flagged"
-        );
-        assert!(
-            !rules.contains("plc-unstructured-jump"),
-            "the CASE state machine uses no JMP"
-        );
-    }
-
-    /// Graphical logic must be analysed too: an FBD POU (blocks + in/out
-    /// variables) is translated to synthetic ST, so the same rules fire on the
-    /// cleartext Modbus block, the hardcoded HMI password and the safety write.
-    #[test]
-    fn fbd_graphical_body_is_analysed() {
-        let all = analyze_tree(&demo_dir(), "demo-target");
-        let fbd: Vec<_> = all
-            .iter()
-            .filter(|f| {
-                f.file_path
-                    .as_deref()
-                    .is_some_and(|p| p.ends_with("pump_fbd.xml"))
-            })
-            .collect();
-        assert!(
-            !fbd.is_empty(),
-            "pump_fbd.xml (FBD) should produce findings"
-        );
-
-        let rules: HashSet<&str> = fbd.iter().filter_map(|f| f.rule_id.as_deref()).collect();
-        for r in [
-            "plc-insecure-comm",          // Modbus_TCP_Master(AUTH := FALSE)
-            "plc-insecure-protocol-port", // PORT := 502
-            "plc-hardcoded-credential",   // HmiPassword := 'admin123'
-            "plc-safety-bypass",          // Safety_Enable := FALSE
-        ] {
-            assert!(
-                rules.contains(r),
-                "expected rule {r} from FBD; got {rules:?}"
-            );
-        }
-    }
-}
@@ -1,766 +0,0 @@
-//! Recursive-descent parser for the security-relevant subset of Structured Text.
-//!
-//! Tolerant by design: it parses the POUs, variable sections, and statement
-//! bodies it understands, and skips (with statement/POU-level recovery) anything
-//! it does not, so a single odd construct never sinks the whole file.
-
-use super::ast::*;
-use super::lexer::{Keyword as K, Tok, Token};
-
-pub struct Parser {
-    toks: Vec<Token>,
-    pos: usize,
-}
-
-impl Parser {
-    pub fn new(toks: Vec<Token>) -> Self {
-        Self { toks, pos: 0 }
-    }
-
-    // ── token helpers ──────────────────────────────────────────────
-    fn peek(&self) -> &Tok {
-        &self.toks[self.pos.min(self.toks.len() - 1)].kind
-    }
-    fn line(&self) -> u32 {
-        self.toks[self.pos.min(self.toks.len() - 1)].line
-    }
-    fn at_end(&self) -> bool {
-        matches!(self.peek(), Tok::Eof)
-    }
-    fn advance(&mut self) -> Tok {
-        let t = self.toks[self.pos.min(self.toks.len() - 1)].kind.clone();
-        if self.pos < self.toks.len() - 1 {
-            self.pos += 1;
-        }
-        t
-    }
-    fn eat(&mut self, t: &Tok) -> bool {
-        if self.peek() == t {
-            self.advance();
-            true
-        } else {
-            false
-        }
-    }
-    fn eat_kw(&mut self, k: K) -> bool {
-        if matches!(self.peek(), Tok::Kw(x) if *x == k) {
-            self.advance();
-            true
-        } else {
-            false
-        }
-    }
-    fn at_kw(&self, k: K) -> bool {
-        matches!(self.peek(), Tok::Kw(x) if *x == k)
-    }
-    fn ident(&mut self) -> Option<String> {
-        if let Tok::Ident(s) = self.peek() {
-            let s = s.clone();
-            self.advance();
-            Some(s)
-        } else {
-            None
-        }
-    }
-
-    // ── top level ──────────────────────────────────────────────────
-    /// Parse every POU in the token stream.
-    pub fn parse_units(&mut self) -> Vec<Pou> {
-        let mut pous = Vec::new();
-        while !self.at_end() {
-            match self.peek() {
-                Tok::Kw(K::Program) => {
-                    self.advance();
-                    if let Some(p) = self.parse_pou(PouKind::Program, K::EndProgram) {
-                        pous.push(p);
-                    }
-                }
-                Tok::Kw(K::Function) => {
-                    self.advance();
-                    if let Some(p) = self.parse_pou(PouKind::Function, K::EndFunction) {
-                        pous.push(p);
-                    }
-                }
-                Tok::Kw(K::FunctionBlock) => {
-                    self.advance();
-                    if let Some(p) = self.parse_pou(PouKind::FunctionBlock, K::EndFunctionBlock) {
-                        pous.push(p);
-                    }
-                }
-                // Skip TYPE...END_TYPE and anything else at top level.
-                _ => {
-                    self.advance();
-                }
-            }
-        }
-        pous
-    }
-
-    fn parse_pou(&mut self, kind: PouKind, end: K) -> Option<Pou> {
-        let line = self.line();
-        let name = self.ident().unwrap_or_else(|| "<anonymous>".to_string());
-        // Optional `: return_type` for functions.
-        if self.eat(&Tok::Colon) {
-            let _ = self.advance(); // return type token
-        }
-
-        let mut vars = Vec::new();
-        // Variable sections precede the body.
-        while let Some(section) = self.var_section_kw() {
-            self.advance();
-            let _ = self.eat_kw(K::Constant); // CONSTANT is informational for our rules
-            self.parse_var_decls(section, &mut vars);
-        }
-
-        // Body statements until END_<kind>.
-        let mut body = Vec::new();
-        while !self.at_end() && !self.at_kw(end) {
-            if let Some(s) = self.parse_stmt() {
-                body.push(s);
-            }
-        }
-        self.eat_kw(end);
-
-        Some(Pou {
-            name,
-            kind,
-            vars,
-            body,
-            line,
-        })
-    }
-
-    fn var_section_kw(&self) -> Option<VarSection> {
-        match self.peek() {
-            Tok::Kw(K::Var) => Some(VarSection::Var),
-            Tok::Kw(K::VarInput) => Some(VarSection::Input),
-            Tok::Kw(K::VarOutput) => Some(VarSection::Output),
-            Tok::Kw(K::VarInOut) => Some(VarSection::InOut),
-            Tok::Kw(K::VarGlobal) => Some(VarSection::Global),
-            Tok::Kw(K::VarTemp) => Some(VarSection::Temp),
-            Tok::Kw(K::VarExternal) => Some(VarSection::External),
-            _ => None,
-        }
-    }
-
-    fn parse_var_decls(&mut self, section: VarSection, out: &mut Vec<VarDecl>) {
-        while !self.at_end() && !self.at_kw(K::EndVar) {
-            let line = self.line();
-            // names: a, b, c
-            let mut names = Vec::new();
-            match self.ident() {
-                Some(n) => names.push(n),
-                None => {
-                    // Not a declaration we understand — skip to next ; or END_VAR.
-                    self.sync_decl();
-                    continue;
-                }
-            }
-            while self.eat(&Tok::Comma) {
-                if let Some(n) = self.ident() {
-                    names.push(n);
-                }
-            }
-            if !self.eat(&Tok::Colon) {
-                self.sync_decl();
-                continue;
-            }
-            let (type_name, array_bounds) = self.parse_type();
-            let init = if self.eat(&Tok::Assign) {
-                Some(self.parse_expr())
-            } else {
-                None
-            };
-            self.eat(&Tok::Semi);
-            for n in names {
-                out.push(VarDecl {
-                    name: n,
-                    section,
-                    type_name: type_name.clone(),
-                    array_bounds,
-                    init: init.clone(),
-                    line,
-                });
-            }
-        }
-        self.eat_kw(K::EndVar);
-    }
-
-    /// Parse a (possibly ARRAY) type, returning its rendered name and literal
-    /// bounds when present.
-    fn parse_type(&mut self) -> (String, Option<(i64, i64)>) {
-        if self.eat_kw(K::Array) {
-            let mut bounds = None;
-            if self.eat(&Tok::LBrack) {
-                let lo = self.int_lit();
-                self.eat(&Tok::DotDot);
-                let hi = self.int_lit();
-                if let (Some(lo), Some(hi)) = (lo, hi) {
-                    bounds = Some((lo, hi));
-                }
-                // Skip any further dimensions / tokens to the closing bracket.
-                while !self.at_end() && !self.eat(&Tok::RBrack) {
-                    self.advance();
-                }
-            }
-            self.eat_kw(K::Of);
-            let elem = self.type_ident();
-            (format!("ARRAY OF {elem}"), bounds)
-        } else {
-            (self.type_ident(), None)
-        }
-    }
-
-    fn type_ident(&mut self) -> String {
-        // Types can be qualified idents; keep it simple: one token, plus any
-        // string-length suffix like STRING[80].
-        let base = match self.advance() {
-            Tok::Ident(s) => s,
-            Tok::Kw(_) => "TYPE".to_string(),
-            other => format!("{other:?}"),
-        };
-        if self.eat(&Tok::LBrack) {
-            while !self.at_end() && !self.eat(&Tok::RBrack) {
-                self.advance();
-            }
-        }
-        base
-    }
-
-    fn int_lit(&mut self) -> Option<i64> {
-        match self.peek() {
-            Tok::Int(n) => {
-                let n = *n;
-                self.advance();
-                Some(n)
-            }
-            Tok::Minus => {
-                self.advance();
-                if let Tok::Int(n) = self.peek() {
-                    let n = -*n;
-                    self.advance();
-                    Some(n)
-                } else {
-                    None
-                }
-            }
-            _ => None,
-        }
-    }
-
-    fn sync_decl(&mut self) {
-        while !self.at_end() && !self.eat(&Tok::Semi) && !self.at_kw(K::EndVar) {
-            self.advance();
-        }
-    }
-    fn sync_stmt(&mut self) {
-        while !self.at_end() && !self.eat(&Tok::Semi) {
-            // Stop at block terminators so recovery doesn't swallow structure.
-            if matches!(
-                self.peek(),
-                Tok::Kw(
-                    K::EndIf
-                        | K::EndFor
-                        | K::EndWhile
-                        | K::EndCase
-                        | K::EndRepeat
-                        | K::EndProgram
-                        | K::EndFunction
-                        | K::EndFunctionBlock
-                        | K::Else
-                        | K::Elsif
-                )
-            ) {
-                return;
-            }
-            self.advance();
-        }
-    }
-
-    // ── statements ─────────────────────────────────────────────────
-    fn parse_stmt(&mut self) -> Option<Stmt> {
-        let line = self.line();
-        match self.peek().clone() {
-            Tok::Semi => {
-                self.advance();
-                None
-            }
-            Tok::Kw(K::If) => self.parse_if(),
-            Tok::Kw(K::Case) => self.parse_case(),
-            Tok::Kw(K::For) => self.parse_for(),
-            Tok::Kw(K::While) => self.parse_while(),
-            Tok::Kw(K::Repeat) => self.parse_repeat(),
-            Tok::Kw(K::Return) => {
-                self.advance();
-                self.eat(&Tok::Semi);
-                Some(Stmt::Return { line })
-            }
-            Tok::Kw(K::Exit) => {
-                self.advance();
-                self.eat(&Tok::Semi);
-                Some(Stmt::Exit { line })
-            }
-            Tok::Kw(K::Jmp) => {
-                self.advance();
-                let label = self.ident().unwrap_or_default();
-                self.eat(&Tok::Semi);
-                Some(Stmt::Jump { label, line })
-            }
-            Tok::Ident(name) => {
-                // Could be `label:`, `call(...)`, or an assignment.
-                // Lookahead: ident ':' (not ':=') → label.
-                if matches!(
-                    self.toks.get(self.pos + 1).map(|t| &t.kind),
-                    Some(Tok::Colon)
-                ) && !matches!(self.toks.get(self.pos + 2).map(|t| &t.kind), Some(Tok::Eq))
-                {
-                    self.advance(); // ident
-                    self.advance(); // ':'
-                    return Some(Stmt::Label { name, line });
-                }
-                let lhs = self.parse_expr();
-                if self.eat(&Tok::Assign) {
-                    let value = self.parse_expr();
-                    self.eat(&Tok::Semi);
-                    Some(Stmt::Assign {
-                        target: lhs,
-                        value,
-                        line,
-                    })
-                } else if let Expr::Call { callee, args, .. } = lhs {
-                    self.eat(&Tok::Semi);
-                    Some(Stmt::Call { callee, args, line })
-                } else {
-                    // Bare expression / FB invocation without args recognized —
-                    // skip to the terminator.
-                    self.sync_stmt();
-                    None
-                }
-            }
-            _ => {
-                self.sync_stmt();
-                None
-            }
-        }
-    }
-
-    fn parse_block_until(&mut self, terms: &[K]) -> Vec<Stmt> {
-        let mut body = Vec::new();
-        while !self.at_end() && !terms.iter().any(|k| self.at_kw(*k)) {
-            if let Some(s) = self.parse_stmt() {
-                body.push(s);
-            }
-        }
-        body
-    }
-
-    fn parse_if(&mut self) -> Option<Stmt> {
-        let line = self.line();
-        self.eat_kw(K::If);
-        let mut branches = Vec::new();
-        let cond = self.parse_expr();
-        self.eat_kw(K::Then);
-        let body = self.parse_block_until(&[K::Elsif, K::Else, K::EndIf]);
-        branches.push((cond, body));
-        while self.eat_kw(K::Elsif) {
-            let c = self.parse_expr();
-            self.eat_kw(K::Then);
-            let b = self.parse_block_until(&[K::Elsif, K::Else, K::EndIf]);
-            branches.push((c, b));
-        }
-        let else_body = if self.eat_kw(K::Else) {
-            Some(self.parse_block_until(&[K::EndIf]))
-        } else {
-            None
-        };
-        self.eat_kw(K::EndIf);
-        self.eat(&Tok::Semi);
-        Some(Stmt::If {
-            branches,
-            else_body,
-            line,
-        })
-    }
-
-    fn parse_case(&mut self) -> Option<Stmt> {
-        let line = self.line();
-        self.eat_kw(K::Case);
-        let selector = self.parse_expr();
-        self.eat_kw(K::Of);
-        let mut arms = Vec::new();
-        let mut else_body = None;
-        while !self.at_end() && !self.at_kw(K::EndCase) {
-            if self.eat_kw(K::Else) {
-                else_body = Some(self.parse_block_until(&[K::EndCase]));
-                break;
-            }
-            // labels: expr {, expr} :
-            let mut labels = vec![self.parse_expr()];
-            while self.eat(&Tok::Comma) {
-                labels.push(self.parse_expr());
-            }
-            self.eat(&Tok::Colon);
-            let body = self.parse_block_until(&[K::EndCase, K::Else]);
-            arms.push((labels, body));
-        }
-        self.eat_kw(K::EndCase);
-        self.eat(&Tok::Semi);
-        Some(Stmt::Case {
-            selector,
-            arms,
-            else_body,
-            line,
-        })
-    }
-
-    fn parse_for(&mut self) -> Option<Stmt> {
-        let line = self.line();
-        self.eat_kw(K::For);
-        let var = self.ident().unwrap_or_default();
-        self.eat(&Tok::Assign);
-        let from = self.parse_expr();
-        self.eat_kw(K::To);
-        let to = self.parse_expr();
-        let by = if self.eat_kw(K::By) {
-            Some(self.parse_expr())
-        } else {
-            None
-        };
-        self.eat_kw(K::Do);
-        let body = self.parse_block_until(&[K::EndFor]);
-        self.eat_kw(K::EndFor);
-        self.eat(&Tok::Semi);
-        Some(Stmt::For {
-            var,
-            from,
-            to,
-            by,
-            body,
-            line,
-        })
-    }
-
-    fn parse_while(&mut self) -> Option<Stmt> {
-        let line = self.line();
-        self.eat_kw(K::While);
-        let cond = self.parse_expr();
-        self.eat_kw(K::Do);
-        let body = self.parse_block_until(&[K::EndWhile]);
-        self.eat_kw(K::EndWhile);
-        self.eat(&Tok::Semi);
-        Some(Stmt::While { cond, body, line })
-    }
-
-    fn parse_repeat(&mut self) -> Option<Stmt> {
-        let line = self.line();
-        self.eat_kw(K::Repeat);
-        let body = self.parse_block_until(&[K::Until, K::EndRepeat]);
-        self.eat_kw(K::Until);
-        let until = self.parse_expr();
-        self.eat_kw(K::EndRepeat);
-        self.eat(&Tok::Semi);
-        Some(Stmt::Repeat { body, until, line })
-    }
-
-    // ── expressions (precedence climbing) ──────────────────────────
-    pub fn parse_expr(&mut self) -> Expr {
-        self.parse_or()
-    }
-
-    fn parse_or(&mut self) -> Expr {
-        let mut lhs = self.parse_and();
-        loop {
-            let op = match self.peek() {
-                Tok::Kw(K::Or) => BinOp::Or,
-                Tok::Kw(K::Xor) => BinOp::Xor,
-                _ => break,
-            };
-            let line = self.line();
-            self.advance();
-            let rhs = self.parse_and();
-            lhs = Expr::Binary {
-                op,
-                lhs: Box::new(lhs),
-                rhs: Box::new(rhs),
-                line,
-            };
-        }
-        lhs
-    }
-
-    fn parse_and(&mut self) -> Expr {
-        let mut lhs = self.parse_cmp();
-        while matches!(self.peek(), Tok::Kw(K::And) | Tok::Amp) {
-            let op = BinOp::And;
-            let line = self.line();
-            self.advance();
-            let rhs = self.parse_cmp();
-            lhs = Expr::Binary {
-                op,
-                lhs: Box::new(lhs),
-                rhs: Box::new(rhs),
-                line,
-            };
-        }
-        lhs
-    }
-
-    fn parse_cmp(&mut self) -> Expr {
-        let mut lhs = self.parse_add();
-        loop {
-            let op = match self.peek() {
-                Tok::Eq => BinOp::Eq,
-                Tok::Ne => BinOp::Ne,
-                Tok::Lt => BinOp::Lt,
-                Tok::Le => BinOp::Le,
-                Tok::Gt => BinOp::Gt,
-                Tok::Ge => BinOp::Ge,
-                _ => break,
-            };
-            let line = self.line();
-            self.advance();
-            let rhs = self.parse_add();
-            lhs = Expr::Binary {
-                op,
-                lhs: Box::new(lhs),
-                rhs: Box::new(rhs),
-                line,
-            };
-        }
-        lhs
-    }
-
-    fn parse_add(&mut self) -> Expr {
-        let mut lhs = self.parse_mul();
-        loop {
-            let op = match self.peek() {
-                Tok::Plus => BinOp::Add,
-                Tok::Minus => BinOp::Sub,
-                _ => break,
-            };
-            let line = self.line();
-            self.advance();
-            let rhs = self.parse_mul();
-            lhs = Expr::Binary {
-                op,
-                lhs: Box::new(lhs),
-                rhs: Box::new(rhs),
-                line,
-            };
-        }
-        lhs
-    }
-
-    fn parse_mul(&mut self) -> Expr {
-        let mut lhs = self.parse_unary();
-        loop {
-            let op = match self.peek() {
-                Tok::Star => BinOp::Mul,
-                Tok::Slash => BinOp::Div,
-                Tok::Kw(K::Mod) => BinOp::Mod,
-                Tok::Power => BinOp::Pow,
-                _ => break,
-            };
-            let line = self.line();
-            self.advance();
-            let rhs = self.parse_unary();
-            lhs = Expr::Binary {
-                op,
-                lhs: Box::new(lhs),
-                rhs: Box::new(rhs),
-                line,
-            };
-        }
-        lhs
-    }
-
-    fn parse_unary(&mut self) -> Expr {
-        let line = self.line();
-        match self.peek() {
-            Tok::Kw(K::Not) => {
-                self.advance();
-                Expr::Unary {
-                    op: UnOp::Not,
-                    expr: Box::new(self.parse_unary()),
-                    line,
-                }
-            }
-            Tok::Minus => {
-                self.advance();
-                Expr::Unary {
-                    op: UnOp::Neg,
-                    expr: Box::new(self.parse_unary()),
-                    line,
-                }
-            }
-            _ => self.parse_postfix(),
-        }
-    }
-
-    fn parse_postfix(&mut self) -> Expr {
-        let mut e = self.parse_primary();
-        loop {
-            let line = self.line();
-            match self.peek() {
-                Tok::LBrack => {
-                    self.advance();
-                    let index = self.parse_expr();
-                    self.eat(&Tok::RBrack);
-                    e = Expr::Index {
-                        base: Box::new(e),
-                        index: Box::new(index),
-                        line,
-                    };
-                }
-                Tok::Dot => {
-                    self.advance();
-                    let field = self.ident().unwrap_or_default();
-                    e = Expr::Member {
-                        base: Box::new(e),
-                        field,
-                        line,
-                    };
-                }
-                _ => break,
-            }
-        }
-        e
-    }
-
-    fn parse_primary(&mut self) -> Expr {
-        let line = self.line();
-        match self.advance() {
-            Tok::Int(n) => Expr::Int(n, line),
-            Tok::Real(r) => Expr::Real(r, line),
-            Tok::Bool(b) => Expr::Bool(b, line),
-            Tok::Str(s) => Expr::Str(s, line),
-            Tok::Time(t) => Expr::Time(t, line),
-            Tok::LParen => {
-                let e = self.parse_expr();
-                self.eat(&Tok::RParen);
-                e
-            }
-            Tok::Ident(name) => {
-                if self.eat(&Tok::LParen) {
-                    let args = self.parse_call_args();
-                    Expr::Call {
-                        callee: name,
-                        args,
-                        line,
-                    }
-                } else {
-                    Expr::Ident(name, line)
-                }
-            }
-            // Unrecognized start of expression — yield a placeholder identifier.
-            _ => Expr::Ident(String::new(), line),
-        }
-    }
-
-    fn parse_call_args(&mut self) -> Vec<CallArg> {
-        let mut args = Vec::new();
-        if self.eat(&Tok::RParen) {
-            return args;
-        }
-        loop {
-            // Named arg: ident := expr  (peek two tokens).
-            if let Tok::Ident(name) = self.peek().clone() {
-                if matches!(
-                    self.toks.get(self.pos + 1).map(|t| &t.kind),
-                    Some(Tok::Assign)
-                ) {
-                    self.advance(); // ident
-                    self.advance(); // :=
-                    let value = self.parse_expr();
-                    args.push(CallArg {
-                        name: Some(name),
-                        value,
-                    });
-                    if self.eat(&Tok::Comma) {
-                        continue;
-                    }
-                    break;
-                }
-            }
-            let value = self.parse_expr();
-            args.push(CallArg { name: None, value });
-            if self.eat(&Tok::Comma) {
-                continue;
-            }
-            break;
-        }
-        self.eat(&Tok::RParen);
-        args
-    }
-}
-
-/// Parse ST source into its POUs.
-pub fn parse(src: &str) -> Vec<Pou> {
-    let toks = super::lexer::lex(src);
-    Parser::new(toks).parse_units()
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    const SAMPLE: &str = r#"
-PROGRAM Main
-VAR
-    idx : INT;
-    pw : STRING := 'admin123';
-    buf : ARRAY[0..9] OF INT;
-    ok : BOOL := FALSE;
-END_VAR
-    // a comment
-    IF idx > 0 THEN
-        buf[idx] := idx * 2;
-    ELSE
-        JMP done;
-    END_IF;
-    Comm(IP := '10.0.0.1', PORT := 502);
-done:
-    ok := TRUE;
-END_PROGRAM
-"#;
-
-    #[test]
-    fn parses_program_vars_and_body() {
-        let pous = parse(SAMPLE);
-        assert_eq!(pous.len(), 1, "one POU");
-        let p = &pous[0];
-        assert_eq!(p.name, "Main");
-        assert_eq!(p.kind, PouKind::Program);
-        // vars: idx, pw, buf, ok
-        assert_eq!(p.vars.len(), 4);
-        let pw = p.vars.iter().find(|v| v.name == "pw").expect("pw");
-        assert!(matches!(&pw.init, Some(Expr::Str(s, _)) if s == "admin123"));
-        let buf = p.vars.iter().find(|v| v.name == "buf").expect("buf");
-        assert_eq!(buf.array_bounds, Some((0, 9)));
-        // body has an IF, a Call, a Label, and an Assign
-        assert!(p.body.iter().any(|s| matches!(s, Stmt::If { .. })));
-        assert!(p
-            .body
-            .iter()
-            .any(|s| matches!(s, Stmt::Call { callee, .. } if callee == "Comm")));
-        assert!(p
-            .body
-            .iter()
-            .any(|s| matches!(s, Stmt::Label { name, .. } if name == "done")));
-    }
-
-    #[test]
-    fn jmp_inside_if_is_captured() {
-        let pous = parse(SAMPLE);
-        let p = &pous[0];
-        // find the IF, check its else branch has a JMP
-        let has_jmp = p.body.iter().any(|s| match s {
-            Stmt::If { else_body, .. } => else_body
-                .as_ref()
-                .map(|b| b.iter().any(|s| matches!(s, Stmt::Jump { .. })))
-                .unwrap_or(false),
-            _ => false,
-        });
-        assert!(has_jmp, "JMP should be parsed inside the ELSE branch");
-    }
-}
@@ -1,418 +0,0 @@
-//! PLCopen XML → Structured Text POUs.
-//!
-//! A PLCopen project stores each POU as `<pou name=".." pouType="..">` with an
-//! `<interface>` (typed variable sections) and a `<body>` in one of the IEC
-//! 61131-3 languages. We reconstruct an equivalent Structured-Text source for
-//! each POU (a `VAR` block from the interface + statements from the body) and run
-//! it through the ST parser, so raw `.st` files and PLCopen projects — textual or
-//! graphical — flow through one analysis path.
-//!
-//! Body languages:
-//! - **ST** — taken verbatim.
-//! - **FBD / LD** — the graphical network is translated to synthetic ST: blocks
-//!   become calls (`TypeName(pin := arg, …)`), out-variables / coils become
-//!   assignments, with input pins resolved by tracing connections. This lets the
-//!   semantic rules see comm calls, hardcoded arguments and safety writes that
-//!   live in graphical logic, not just in text.
-//! - **SFC** — the step/transition graph itself is skipped; the ST/FBD/LD bodies
-//!   embedded in its actions and transitions are still translated.
-
-use std::collections::HashMap;
-
-use roxmltree::Node;
-
-use super::ast::Pou;
-use super::parser;
-
-/// Parse every POU out of a PLCopen XML document (ST, FBD or LD bodies).
-pub fn parse_plcopen(xml: &str) -> Vec<Pou> {
-    let doc = match roxmltree::Document::parse(xml) {
-        Ok(d) => d,
-        Err(_) => return Vec::new(),
-    };
-    let mut pous = Vec::new();
-    for pou in doc.descendants().filter(|n| n.has_tag_name("pou")) {
-        let name = pou.attribute("name").unwrap_or("pou").to_string();
-        let pou_type = pou.attribute("pouType").unwrap_or("program");
-
-        let Some(body) = reconstruct_body(pou) else {
-            continue;
-        };
-        if body.trim().is_empty() {
-            continue;
-        }
-
-        let var_block = build_var_block(pou);
-        let kw = match pou_type.to_ascii_lowercase().as_str() {
-            "function" => "FUNCTION",
-            "functionblock" | "functionblocktype" => "FUNCTION_BLOCK",
-            _ => "PROGRAM",
-        };
-        let synthetic = format!("{kw} {name}\n{var_block}{body}\nEND_{kw}\n");
-        pous.extend(parser::parse(&synthetic));
-    }
-    pous
-}
-
-/// Case-insensitive tag match (PLCopen uses `FBD`/`LD`/`ST`, CODESYS may vary).
-fn tag_is(n: &Node, name: &str) -> bool {
-    n.tag_name().name().eq_ignore_ascii_case(name)
-}
-
-/// Reconstruct a POU's body as Structured Text, whatever language it is written
-/// in. Concatenates every language body found under `<body>` (SFC actions and
-/// transitions carry their own ST/FBD/LD sub-bodies).
-fn reconstruct_body(pou: Node) -> Option<String> {
-    let mut out = String::new();
-    for body in pou.descendants().filter(|n| tag_is(n, "body")) {
-        for lang in body.children().filter(|n| n.is_element()) {
-            let piece = match lang.tag_name().name().to_ascii_uppercase().as_str() {
-                "ST" | "IL" => collect_text(lang),
-                "FBD" | "LD" => translate_network(lang),
-                _ => continue,
-            };
-            if !piece.trim().is_empty() {
-                out.push_str(&piece);
-                if !piece.ends_with('\n') {
-                    out.push('\n');
-                }
-            }
-        }
-    }
-    if out.trim().is_empty() {
-        None
-    } else {
-        Some(out)
-    }
-}
-
-// ── graphical (FBD / LD) → synthetic ST ────────────────────────────────
-
-/// Translate one FBD/LD network into ST statements: blocks → calls,
-/// out-variables and coils → assignments.
-fn translate_network(net: Node) -> String {
-    let by_id = index_local_ids(net);
-    let mut out = String::new();
-    for el in net.children().filter(|n| n.is_element()) {
-        let stmt = match el.tag_name().name().to_ascii_lowercase().as_str() {
-            "block" => block_call(el, &by_id).map(|c| format!("{c};")),
-            "outvariable" => out_assignment(el, &by_id),
-            "coil" => coil_assignment(el, &by_id),
-            _ => None,
-        };
-        if let Some(s) = stmt {
-            out.push_str(&s);
-            out.push('\n');
-        }
-    }
-    out
-}
-
-/// Index every element in a network by its `localId` so connections resolve.
-fn index_local_ids<'a, 'input>(net: Node<'a, 'input>) -> HashMap<String, Node<'a, 'input>> {
-    net.descendants()
-        .filter(|n| n.is_element())
-        .filter_map(|n| n.attribute("localId").map(|id| (id.to_string(), n)))
-        .collect()
-}
-
-/// Build a call expression for a block: `TypeName(pin := arg, …)`.
-fn block_call(block: Node, by_id: &HashMap<String, Node>) -> Option<String> {
-    let ty = block.attribute("typeName")?;
-    let mut args = Vec::new();
-    if let Some(inputs) = block.children().find(|n| tag_is(n, "inputVariables")) {
-        for v in inputs.children().filter(|n| tag_is(n, "variable")) {
-            let Some(expr) = input_expr(v, by_id, 0) else {
-                continue;
-            };
-            match v.attribute("formalParameter") {
-                Some(pin) if !pin.is_empty() => args.push(format!("{pin} := {expr}")),
-                _ => args.push(expr),
-            }
-        }
-    }
-    Some(format!("{ty}({})", args.join(", ")))
-}
-
-/// `target := <traced expression>;` for an FBD out-variable.
-fn out_assignment(outvar: Node, by_id: &HashMap<String, Node>) -> Option<String> {
-    let target = expression_text(outvar)?;
-    let value = input_expr(outvar, by_id, 0).unwrap_or_else(|| "0".to_string());
-    Some(format!("{target} := {value};"))
-}
-
-/// `coil := <traced rung expression>;` for an LD coil (negated → `NOT (…)`).
-fn coil_assignment(coil: Node, by_id: &HashMap<String, Node>) -> Option<String> {
-    let target = child_text(coil, "variable")?;
-    let rung = input_expr(coil, by_id, 0).unwrap_or_else(|| "TRUE".to_string());
-    let negated = matches!(coil.attribute("negated"), Some(v) if v.eq_ignore_ascii_case("true"));
-    let rhs = if negated {
-        format!("NOT ({rung})")
-    } else {
-        rung
-    };
-    Some(format!("{target} := {rhs};"))
-}
-
-/// Resolve the expression feeding `node`'s single input connection.
-fn input_expr(node: Node, by_id: &HashMap<String, Node>, depth: u8) -> Option<String> {
-    let refid = ref_local_id(node)?;
-    Some(expr_for(&refid, by_id, depth))
-}
-
-/// Build the ST expression produced by the element with this `localId`.
-fn expr_for(local_id: &str, by_id: &HashMap<String, Node>, depth: u8) -> String {
-    if depth > 24 {
-        return "0".to_string();
-    }
-    let Some(node) = by_id.get(local_id) else {
-        return format!("__net{local_id}");
-    };
-    match node.tag_name().name().to_ascii_lowercase().as_str() {
-        "invariable" | "inoutvariable" => {
-            expression_text(*node).unwrap_or_else(|| format!("__net{local_id}"))
-        }
-        // A block feeding another element: reference it by a synthetic result
-        // name; the block is emitted as its own call statement, so we neither
-        // duplicate the call nor lose it.
-        "block" => format!("__blk{local_id}"),
-        "contact" => {
-            let var = child_text(*node, "variable").unwrap_or_else(|| "TRUE".to_string());
-            let negated =
-                matches!(node.attribute("negated"), Some(v) if v.eq_ignore_ascii_case("true"));
-            let term = if negated { format!("NOT {var}") } else { var };
-            match ref_local_id(*node) {
-                Some(up) => {
-                    let upstream = expr_for(&up, by_id, depth + 1);
-                    if upstream == "TRUE" {
-                        term
-                    } else {
-                        format!("({upstream} AND {term})")
-                    }
-                }
-                None => term,
-            }
-        }
-        "leftpowerrail" => "TRUE".to_string(),
-        _ => format!("__net{local_id}"),
-    }
-}
-
-/// The `refLocalId` of `node`'s first input connection, if any.
-fn ref_local_id(node: Node) -> Option<String> {
-    node.descendants()
-        .find(|n| tag_is(n, "connectionPointIn"))
-        .and_then(|cpi| cpi.descendants().find(|n| tag_is(n, "connection")))
-        .and_then(|c| c.attribute("refLocalId"))
-        .map(|s| s.to_string())
-}
-
-/// Text of a node's `<expression>` child (variable name or literal).
-fn expression_text(node: Node) -> Option<String> {
-    let e = node.children().find(|n| tag_is(n, "expression"))?;
-    let t = collect_text(e).trim().to_string();
-    if t.is_empty() {
-        None
-    } else {
-        Some(t)
-    }
-}
-
-/// Text of a named child element (e.g. `<variable>` of a contact/coil).
-fn child_text(node: Node, name: &str) -> Option<String> {
-    let c = node.children().find(|n| tag_is(n, name))?;
-    let t = collect_text(c).trim().to_string();
-    if t.is_empty() {
-        None
-    } else {
-        Some(t)
-    }
-}
-
-/// Concatenate the text of a node's descendant text nodes (bodies are often
-/// wrapped in `<xhtml>` and may contain multiple text runs). Only text nodes are
-/// gathered: an element's `.text()` would re-yield its first child's text, which
-/// (with the text node itself) would duplicate every value.
-fn collect_text(node: Node) -> String {
-    node.descendants()
-        .filter(|n| n.is_text())
-        .filter_map(|n| n.text())
-        .collect::<String>()
-}
-
-/// Build an ST `VAR … END_VAR` block from a POU's `<interface>` variable
-/// sections, so declarations (types, initial values) reach the rules.
-fn build_var_block(pou: Node) -> String {
-    let Some(interface) = pou.children().find(|n| n.has_tag_name("interface")) else {
-        return String::new();
-    };
-    let mut out = String::from("VAR\n");
-    let mut any = false;
-    for container in interface.children().filter(|n| n.is_element()) {
-        // localVars / inputVars / outputVars / inOutVars / tempVars / globalVars / externalVars
-        if !container.tag_name().name().ends_with("Vars") {
-            continue;
-        }
-        for var in container.children().filter(|n| n.has_tag_name("variable")) {
-            let Some(vname) = var.attribute("name") else {
-                continue;
-            };
-            let ty = var
-                .children()
-                .find(|n| n.has_tag_name("type"))
-                .map(type_name)
-                .unwrap_or_else(|| "BOOL".to_string());
-            let init = var
-                .children()
-                .find(|n| n.has_tag_name("initialValue"))
-                .and_then(initial_value);
-            match init {
-                Some(v) => out.push_str(&format!("  {vname} : {ty} := {v};\n")),
-                None => out.push_str(&format!("  {vname} : {ty};\n")),
-            }
-            any = true;
-        }
-    }
-    out.push_str("END_VAR\n");
-    if any {
-        out
-    } else {
-        String::new()
-    }
-}
-
-/// Render a PLCopen `<type>` element as an ST type string.
-fn type_name(type_node: Node) -> String {
-    let Some(inner) = type_node.children().find(|n| n.is_element()) else {
-        return "BOOL".to_string();
-    };
-    let tag = inner.tag_name().name();
-    match tag {
-        "derived" => inner.attribute("name").unwrap_or("DERIVED").to_string(),
-        "array" => {
-            let dim = inner.children().find(|n| n.has_tag_name("dimension"));
-            let (lo, hi) = dim
-                .map(|d| {
-                    (
-                        d.attribute("lower").unwrap_or("0").to_string(),
-                        d.attribute("upper").unwrap_or("0").to_string(),
-                    )
-                })
-                .unwrap_or_else(|| ("0".to_string(), "0".to_string()));
-            let base = inner
-                .children()
-                .find(|n| n.has_tag_name("baseType"))
-                .map(type_name)
-                .unwrap_or_else(|| "INT".to_string());
-            format!("ARRAY[{lo}..{hi}] OF {base}")
-        }
-        "string" | "wstring" => "STRING".to_string(),
-        // BOOL, INT, DINT, REAL, TIME, ... — the tag name is the ST type.
-        other => other.to_ascii_uppercase(),
-    }
-}
-
-/// Extract an initial value as an ST literal (quoting strings).
-fn initial_value(iv: Node) -> Option<String> {
-    let simple = iv.descendants().find(|n| n.has_tag_name("simpleValue"))?;
-    let raw = simple.attribute("value")?.trim().to_string();
-    if raw.is_empty() {
-        return None;
-    }
-    // Numbers / booleans / time literals pass through; everything else is a
-    // string literal.
-    let is_scalar = raw.eq_ignore_ascii_case("true")
-        || raw.eq_ignore_ascii_case("false")
-        || raw.starts_with(['T', 't', 'D', 'd']) && raw.contains('#')
-        || raw
-            .chars()
-            .all(|c| c.is_ascii_digit() || c == '.' || c == '-' || c == '+');
-    if is_scalar || raw.starts_with('\'') || raw.starts_with('"') {
-        Some(raw)
-    } else {
-        Some(format!("'{}'", raw.replace('\'', "''")))
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use super::parse_plcopen;
-    use crate::pipeline::plc::rules;
-    use std::collections::HashSet;
-
-    fn rule_ids(xml: &str) -> HashSet<&'static str> {
-        parse_plcopen(xml)
-            .iter()
-            .flat_map(rules::analyze)
-            .map(|h| h.rule_id)
-            .collect()
-    }
-
-    /// A Ladder Diagram network: a rung (power rail → contact → coil) plus an
-    /// insecure comm block. Coils/contacts translate to assignments; the block
-    /// translates to a call so the port rule fires.
-    #[test]
-    fn ld_coil_and_block_translate_and_are_analysed() {
-        let xml = r#"<?xml version="1.0"?>
-<project xmlns="http://www.plcopen.org/xml/tc6_0201">
- <types><pous>
-  <pou name="Rung" pouType="program">
-   <interface><localVars>
-     <variable name="Motor"><type><BOOL/></type></variable>
-   </localVars></interface>
-   <body><LD>
-     <leftPowerRail localId="0"/>
-     <contact localId="1"><variable>Start</variable>
-       <connectionPointIn><connection refLocalId="0"/></connectionPointIn></contact>
-     <coil localId="2"><variable>Motor</variable>
-       <connectionPointIn><connection refLocalId="1"/></connectionPointIn></coil>
-     <inVariable localId="3"><expression>21</expression></inVariable>
-     <inVariable localId="4"><expression>FALSE</expression></inVariable>
-     <block localId="10" typeName="Ftp_Send">
-       <inputVariables>
-         <variable formalParameter="PORT">
-           <connectionPointIn><connection refLocalId="3"/></connectionPointIn></variable>
-         <variable formalParameter="ENCRYPT">
-           <connectionPointIn><connection refLocalId="4"/></connectionPointIn></variable>
-       </inputVariables>
-     </block>
-   </LD></body>
-  </pou>
- </pous></types>
-</project>"#;
-        let ids = rule_ids(xml);
-        // Ftp_Send(PORT := 21, ENCRYPT := FALSE) — port 21 is an insecure protocol.
-        assert!(
-            ids.contains("plc-insecure-protocol-port"),
-            "LD block should flag port 21; got {ids:?}"
-        );
-    }
-
-    /// Doubled-text regression: a graphical expression must be extracted once,
-    /// so literals like `502` and `FALSE` stay intact (not `502502`/`FALSEFALSE`).
-    #[test]
-    fn graphical_expression_text_is_not_duplicated() {
-        let xml = r#"<?xml version="1.0"?>
-<project xmlns="http://www.plcopen.org/xml/tc6_0201">
- <types><pous>
-  <pou name="Comm" pouType="program">
-   <body><FBD>
-     <inVariable localId="1"><expression>502</expression></inVariable>
-     <inVariable localId="2"><expression>FALSE</expression></inVariable>
-     <block localId="10" typeName="Modbus_TCP_Master">
-       <inputVariables>
-         <variable formalParameter="PORT">
-           <connectionPointIn><connection refLocalId="1"/></connectionPointIn></variable>
-         <variable formalParameter="AUTH">
-           <connectionPointIn><connection refLocalId="2"/></connectionPointIn></variable>
-       </inputVariables>
-     </block>
-   </FBD></body>
-  </pou>
- </pous></types>
-</project>"#;
-        let ids = rule_ids(xml);
-        assert!(ids.contains("plc-insecure-protocol-port")); // PORT := 502 (not 502502)
-        assert!(ids.contains("plc-insecure-comm")); // AUTH := FALSE (not FALSEFALSE)
-    }
-}
@@ -1,632 +0,0 @@
-//! Semantic control-logic security rules over the Structured Text AST.
-//!
-//! Each rule walks the parsed [`Pou`] and yields [`RuleHit`]s the scanner turns
-//! into findings. Rules reason over structure (declarations, assignments, calls,
-//! array accesses, division, jumps) rather than raw text, so they see through
-//! formatting and comments.
-
-use std::collections::{HashMap, HashSet};
-
-use compliance_core::models::Severity;
-
-use super::ast::*;
-
-/// One rule match within a POU.
-pub struct RuleHit {
-    pub line: u32,
-    pub severity: Severity,
-    pub rule_id: &'static str,
-    pub title: String,
-    pub description: String,
-    pub cwe: Option<&'static str>,
-    pub remediation: &'static str,
-}
-
-/// Run every rule over a POU.
-pub fn analyze(pou: &Pou) -> Vec<RuleHit> {
-    let mut hits = Vec::new();
-    let ctx = Ctx::build(pou);
-
-    // Declaration-level rules.
-    for v in &pou.vars {
-        if let Some(init) = &v.init {
-            check_credential_binding(&v.name, init, &pou.name, &mut hits);
-            check_default_password(init, &v.name, &pou.name, &mut hits);
-        }
-    }
-
-    // Body walk.
-    walk(&pou.body, pou, &ctx, &GuardSet::default(), &mut hits);
-    hits
-}
-
-/// Per-POU context precomputed once.
-struct Ctx {
-    /// Names declared in VAR_INPUT (untrusted / externally driven).
-    input_vars: HashSet<String>,
-    /// Array variable name → declared (lo, hi) bounds.
-    arrays: HashMap<String, (i64, i64)>,
-}
-
-impl Ctx {
-    fn build(pou: &Pou) -> Self {
-        let mut input_vars = HashSet::new();
-        let mut arrays = HashMap::new();
-        for v in &pou.vars {
-            if v.section == VarSection::Input {
-                input_vars.insert(v.name.to_ascii_lowercase());
-            }
-            if let Some(b) = v.array_bounds {
-                arrays.insert(v.name.to_ascii_lowercase(), b);
-            }
-        }
-        Self { input_vars, arrays }
-    }
-}
-
-/// Variables proven non-zero on the current control-flow path (from enclosing
-/// `IF`/`WHILE` conditions), so guarded divisions aren't false-flagged.
-#[derive(Default, Clone)]
-struct GuardSet {
-    nonzero: HashSet<String>,
-}
-
-impl GuardSet {
-    fn with(&self, names: Vec<String>) -> Self {
-        let mut g = self.clone();
-        g.nonzero.extend(names);
-        g
-    }
-    fn is_nonzero(&self, name: &str) -> bool {
-        self.nonzero.contains(name)
-    }
-}
-
-/// Variable names a condition proves non-zero (`v <> 0`, `v > 0`, `v >= 1`,
-/// `v < 0`, and conjunctions thereof).
-fn guards_from_cond(cond: &Expr) -> Vec<String> {
-    let mut out = Vec::new();
-    collect_nonzero(cond, &mut out);
-    out
-}
-
-fn collect_nonzero(e: &Expr, out: &mut Vec<String>) {
-    let Expr::Binary { op, lhs, rhs, .. } = e else {
-        return;
-    };
-    let is_zero = |x: &Expr| {
-        matches!(x, Expr::Int(0, _)) || matches!(x, Expr::Real(r, _) if r.abs() < f64::EPSILON)
-    };
-    let int_of = |x: &Expr| match x {
-        Expr::Int(n, _) => Some(*n),
-        _ => None,
-    };
-    match op {
-        BinOp::And => {
-            collect_nonzero(lhs, out);
-            collect_nonzero(rhs, out);
-        }
-        BinOp::Ne => {
-            if let (Some(v), true) = (lhs.as_ident(), is_zero(rhs)) {
-                out.push(v.to_ascii_lowercase());
-            }
-            if let (true, Some(v)) = (is_zero(lhs), rhs.as_ident()) {
-                out.push(v.to_ascii_lowercase());
-            }
-        }
-        BinOp::Gt | BinOp::Lt => {
-            // v > 0  or  v < 0
-            if let (Some(v), true) = (lhs.as_ident(), is_zero(rhs)) {
-                out.push(v.to_ascii_lowercase());
-            }
-        }
-        BinOp::Ge => {
-            // v >= n, n >= 1
-            if let (Some(v), Some(n)) = (lhs.as_ident(), int_of(rhs)) {
-                if n >= 1 {
-                    out.push(v.to_ascii_lowercase());
-                }
-            }
-        }
-        _ => {}
-    }
-}
-
-// ── the walker ─────────────────────────────────────────────────────
-fn walk(stmts: &[Stmt], pou: &Pou, ctx: &Ctx, guards: &GuardSet, hits: &mut Vec<RuleHit>) {
-    for s in stmts {
-        match s {
-            Stmt::Assign {
-                target,
-                value,
-                line,
-            } => {
-                check_safety_bypass(target, value, *line, &pou.name, hits);
-                // A string bound to a secret-looking target is a credential.
-                if let Some(name) = flatten_ident(target) {
-                    check_credential_binding(&name, value, &pou.name, hits);
-                    check_default_password(value, &name, &pou.name, hits);
-                }
-                walk_expr(target, pou, ctx, guards, hits);
-                walk_expr(value, pou, ctx, guards, hits);
-            }
-            Stmt::Call { callee, args, line } => {
-                check_insecure_comm(callee, args, *line, &pou.name, hits);
-                check_credentials_in_call(callee, args, *line, &pou.name, hits);
-                for a in args {
-                    walk_expr(&a.value, pou, ctx, guards, hits);
-                }
-            }
-            Stmt::Jump { label, line } => hits.push(RuleHit {
-                line: *line,
-                severity: Severity::Medium,
-                rule_id: "plc-unstructured-jump",
-                title: "Unstructured jump (JMP) in control logic".to_string(),
-                description: format!(
-                    "POU `{}` uses `JMP {label}`. Unstructured jumps make control flow hard to \
-                     verify and can bypass safety interlocks or leave outputs in an undefined \
-                     state on unexpected paths.",
-                    pou.name
-                ),
-                cwe: Some("CWE-691"),
-                remediation: "Replace JMP with structured constructs (IF/CASE/loops); reserve \
-                              jumps for well-reviewed state machines only.",
-            }),
-            Stmt::If {
-                branches,
-                else_body,
-                ..
-            } => {
-                for (cond, body) in branches {
-                    walk_expr(cond, pou, ctx, guards, hits);
-                    let child = guards.with(guards_from_cond(cond));
-                    walk(body, pou, ctx, &child, hits);
-                }
-                if let Some(b) = else_body {
-                    walk(b, pou, ctx, guards, hits);
-                }
-            }
-            Stmt::Case {
-                selector,
-                arms,
-                else_body,
-                ..
-            } => {
-                walk_expr(selector, pou, ctx, guards, hits);
-                for (labels, body) in arms {
-                    for l in labels {
-                        walk_expr(l, pou, ctx, guards, hits);
-                    }
-                    walk(body, pou, ctx, guards, hits);
-                }
-                if let Some(b) = else_body {
-                    walk(b, pou, ctx, guards, hits);
-                }
-            }
-            Stmt::For {
-                from, to, by, body, ..
-            } => {
-                walk_expr(from, pou, ctx, guards, hits);
-                walk_expr(to, pou, ctx, guards, hits);
-                if let Some(b) = by {
-                    walk_expr(b, pou, ctx, guards, hits);
-                }
-                walk(body, pou, ctx, guards, hits);
-            }
-            Stmt::While { cond, body, .. } => {
-                walk_expr(cond, pou, ctx, guards, hits);
-                let child = guards.with(guards_from_cond(cond));
-                walk(body, pou, ctx, &child, hits);
-            }
-            Stmt::Repeat { body, until, .. } => {
-                walk(body, pou, ctx, guards, hits);
-                walk_expr(until, pou, ctx, guards, hits);
-            }
-            Stmt::Return { .. } | Stmt::Exit { .. } | Stmt::Label { .. } => {}
-        }
-    }
-}
-
-fn walk_expr(e: &Expr, pou: &Pou, ctx: &Ctx, guards: &GuardSet, hits: &mut Vec<RuleHit>) {
-    match e {
-        Expr::Index { base, index, line } => {
-            check_array_bounds(base, index, *line, ctx, &pou.name, hits);
-            walk_expr(base, pou, ctx, guards, hits);
-            walk_expr(index, pou, ctx, guards, hits);
-        }
-        Expr::Binary { op, lhs, rhs, line } => {
-            if matches!(op, BinOp::Div | BinOp::Mod) {
-                check_division(rhs, *line, &pou.name, guards, hits);
-            }
-            walk_expr(lhs, pou, ctx, guards, hits);
-            walk_expr(rhs, pou, ctx, guards, hits);
-        }
-        Expr::Unary { expr, .. } => walk_expr(expr, pou, ctx, guards, hits),
-        Expr::Member { base, .. } => walk_expr(base, pou, ctx, guards, hits),
-        Expr::Call { args, .. } => {
-            for a in args {
-                walk_expr(&a.value, pou, ctx, guards, hits);
-            }
-        }
-        _ => {}
-    }
-}
-
-// ── individual rules ───────────────────────────────────────────────
-const SECRET_HINTS: &[&str] = &[
-    "password",
-    "passwd",
-    "pwd",
-    "secret",
-    "apikey",
-    "api_key",
-    "token",
-    "credential",
-    "privkey",
-    "private_key",
-    "passphrase",
-];
-
-const DEFAULT_PASSWORDS: &[&str] = &[
-    "admin",
-    "administrator",
-    "password",
-    "passwd",
-    "1234",
-    "12345",
-    "123456",
-    "0000",
-    "1111",
-    "root",
-    "default",
-    "admin123",
-    "changeme",
-    "letmein",
-    "guest",
-    "user",
-    "system",
-    "plc",
-    "codesys",
-];
-
-const COMM_FB_HINTS: &[&str] = &[
-    "modbus", "tcp", "udp", "socket", "mqtt", "opcua", "opc_ua", "ethernet", "ethip", "enip",
-    "dnp3", "ftp", "telnet", "http", "send", "connect", "sock", "comm", "profinet", "s7",
-];
-
-/// Insecure cleartext service ports.
-const INSECURE_PORTS: &[i64] = &[21, 23, 80, 502, 20000, 44818, 102];
-
-fn check_credential_binding(var_name: &str, value: &Expr, pou: &str, hits: &mut Vec<RuleHit>) {
-    let name = var_name.to_ascii_lowercase();
-    let looks_secret = SECRET_HINTS.iter().any(|h| name.contains(h));
-    if looks_secret {
-        if let Expr::Str(s, line) = value {
-            if !s.is_empty() {
-                hits.push(RuleHit {
-                    line: *line,
-                    severity: Severity::High,
-                    rule_id: "plc-hardcoded-credential",
-                    title: "Hardcoded credential in PLC program".to_string(),
-                    description: format!(
-                        "POU `{pou}` binds a hardcoded secret to `{var_name}`. Credentials \
-                         embedded in control logic are extracted trivially from a project export \
-                         or a firmware dump and cannot be rotated without a redeploy."
-                    ),
-                    cwe: Some("CWE-798"),
-                    remediation: "Store secrets outside the program (secure parameter store / \
-                                  operator-entered, retained-but-protected memory); never commit \
-                                  them to the POU.",
-                });
-            }
-        }
-    }
-}
-
-fn check_default_password(value: &Expr, var_name: &str, pou: &str, hits: &mut Vec<RuleHit>) {
-    if let Expr::Str(s, line) = value {
-        let lower = s.to_ascii_lowercase();
-        if DEFAULT_PASSWORDS.contains(&lower.as_str()) {
-            hits.push(RuleHit {
-                line: *line,
-                severity: Severity::Critical,
-                rule_id: "plc-default-password",
-                title: "Default/weak password in PLC program".to_string(),
-                description: format!(
-                    "POU `{pou}` uses the well-known default/weak password `{s}` (bound to \
-                     `{var_name}`). Default PLC credentials are the first thing an attacker tries."
-                ),
-                cwe: Some("CWE-1393"),
-                remediation:
-                    "Require a strong, unique, operator-set password; block commissioning \
-                              until the default is changed.",
-            });
-        }
-    }
-}
-
-fn check_credentials_in_call(
-    callee: &str,
-    args: &[CallArg],
-    line: u32,
-    pou: &str,
-    hits: &mut Vec<RuleHit>,
-) {
-    for a in args {
-        if let Some(name) = &a.name {
-            let n = name.to_ascii_lowercase();
-            if SECRET_HINTS.iter().any(|h| n.contains(h)) {
-                if let Expr::Str(s, l) = &a.value {
-                    if !s.is_empty() {
-                        hits.push(RuleHit {
-                            line: *l,
-                            severity: Severity::High,
-                            rule_id: "plc-hardcoded-credential",
-                            title: "Hardcoded credential passed to a function block".to_string(),
-                            description: format!(
-                                "POU `{pou}` passes a hardcoded secret as `{name}` to `{callee}`."
-                            ),
-                            cwe: Some("CWE-798"),
-                            remediation: "Supply credentials from protected configuration at \
-                                          runtime, not as a literal argument.",
-                        });
-                    }
-                }
-            }
-        }
-    }
-    let _ = line;
-}
-
-fn check_safety_bypass(target: &Expr, value: &Expr, line: u32, pou: &str, hits: &mut Vec<RuleHit>) {
-    let Some(name) = flatten_ident(target) else {
-        return;
-    };
-    let n = name.to_ascii_lowercase();
-    let safety = [
-        "safety",
-        "estop",
-        "e_stop",
-        "emergency",
-        "interlock",
-        "guard",
-        "permit",
-    ]
-    .iter()
-    .any(|h| n.contains(h));
-    let watchdog = n.contains("watchdog") || n.contains("wdt");
-    // A safety enable / interlock / watchdog signal driven to FALSE or 0 in
-    // application logic is a bypass (e.g. `Safety_Enable := FALSE`, `Watchdog_Kick := 0`).
-    let disabling = matches!(value, Expr::Bool(false, _)) || matches!(value, Expr::Int(0, _));
-    if (safety || watchdog) && disabling {
-        hits.push(RuleHit {
-            line,
-            severity: Severity::Critical,
-            rule_id: "plc-safety-bypass",
-            title: "Safety interlock / watchdog disabled in logic".to_string(),
-            description: format!(
-                "POU `{pou}` disables a safety-related signal (`{name}`) in program logic. \
-                 Bypassing interlocks or watchdogs in code defeats the plant's protective \
-                 functions and is a direct hazard."
-            ),
-            cwe: Some("CWE-1384"),
-            remediation: "Never disable safety functions from application logic; safety must be \
-                          handled by a certified safety controller / hard-wired circuit.",
-        });
-    }
-}
-
-fn check_array_bounds(
-    base: &Expr,
-    index: &Expr,
-    line: u32,
-    ctx: &Ctx,
-    pou: &str,
-    hits: &mut Vec<RuleHit>,
-) {
-    // Only reason about arrays we know the bounds of.
-    let Some(arr_name) = base.as_ident() else {
-        return;
-    };
-    if !ctx.arrays.contains_key(&arr_name.to_ascii_lowercase()) {
-        return;
-    }
-    // Index by an untrusted input variable → potential out-of-bounds access.
-    if let Some(idx_name) = index.as_ident() {
-        if ctx.input_vars.contains(&idx_name.to_ascii_lowercase()) {
-            hits.push(RuleHit {
-                line,
-                severity: Severity::High,
-                rule_id: "plc-array-unchecked-index",
-                title: "Array indexed by unvalidated input".to_string(),
-                description: format!(
-                    "POU `{pou}` indexes array `{arr_name}` with the input variable `{idx_name}` \
-                     without a validated bounds check. An out-of-range index corrupts adjacent \
-                     memory or faults the PLC (loss of control)."
-                ),
-                cwe: Some("CWE-129"),
-                remediation: "Clamp or validate the index against the array bounds (e.g. \
-                              `LIMIT`/explicit `IF idx >= lo AND idx <= hi`) before the access.",
-            });
-        }
-    }
-}
-
-fn check_division(
-    divisor: &Expr,
-    line: u32,
-    pou: &str,
-    guards: &GuardSet,
-    hits: &mut Vec<RuleHit>,
-) {
-    // A divisor proven non-zero by an enclosing guard is safe.
-    if let Expr::Ident(name, _) = divisor {
-        if guards.is_nonzero(&name.to_ascii_lowercase()) {
-            return;
-        }
-    }
-    // Flag division by a variable (could be zero); nonzero literals are fine.
-    let risky = matches!(
-        divisor,
-        Expr::Ident(_, _) | Expr::Member { .. } | Expr::Index { .. } | Expr::Int(0, _)
-    );
-    if risky {
-        hits.push(RuleHit {
-            line,
-            severity: Severity::Medium,
-            rule_id: "plc-division-by-zero",
-            title: "Division by a variable without a zero-guard".to_string(),
-            description: format!(
-                "POU `{pou}` divides by a variable that is not proven non-zero. A zero divisor \
-                 raises a PLC exception and can halt the scan cycle (denial of control)."
-            ),
-            cwe: Some("CWE-369"),
-            remediation: "Guard the divisor (`IF d <> 0 THEN …`) or use a safe-divide helper that \
-                          returns a defined value for a zero denominator.",
-        });
-    }
-}
-
-fn check_insecure_comm(
-    callee: &str,
-    args: &[CallArg],
-    line: u32,
-    pou: &str,
-    hits: &mut Vec<RuleHit>,
-) {
-    let c = callee.to_ascii_lowercase();
-    let is_comm = COMM_FB_HINTS.iter().any(|h| c.contains(h));
-    if !is_comm {
-        return;
-    }
-    // Auth/encryption explicitly disabled.
-    for a in args {
-        if let Some(name) = &a.name {
-            let n = name.to_ascii_lowercase();
-            let security_flag = ["auth", "secure", "encrypt", "tls", "ssl", "authentication"]
-                .iter()
-                .any(|h| n.contains(h));
-            if security_flag && matches!(a.value, Expr::Bool(false, _)) {
-                hits.push(RuleHit {
-                    line,
-                    severity: Severity::High,
-                    rule_id: "plc-insecure-comm",
-                    title: "Network communication with security disabled".to_string(),
-                    description: format!(
-                        "POU `{pou}` calls `{callee}` with `{name} := FALSE`, disabling \
-                         authentication/encryption on an industrial network link."
-                    ),
-                    cwe: Some("CWE-319"),
-                    remediation: "Enable authentication + transport encryption; segment OT \
-                                  networks and restrict the endpoint to trusted peers.",
-                });
-            }
-        }
-        // Well-known cleartext port literal.
-        if let Expr::Int(p, _) = &a.value {
-            if INSECURE_PORTS.contains(p) {
-                hits.push(RuleHit {
-                    line,
-                    severity: Severity::Medium,
-                    rule_id: "plc-insecure-protocol-port",
-                    title: "Cleartext industrial protocol port".to_string(),
-                    description: format!(
-                        "POU `{pou}` opens `{callee}` on port {p}, a well-known cleartext OT \
-                         protocol port with no built-in authentication or encryption."
-                    ),
-                    cwe: Some("CWE-319"),
-                    remediation: "Front the protocol with a secure gateway/VPN, or use the \
-                                  authenticated/encrypted variant; never expose it to untrusted \
-                                  networks.",
-                });
-            }
-        }
-    }
-    let _ = line;
-}
-
-/// The dotted/base identifier of an lvalue expression (`a`, `a.b` → `a.b`,
-/// `a[i]` → `a`), for name-based rules.
-fn flatten_ident(e: &Expr) -> Option<String> {
-    match e {
-        Expr::Ident(n, _) => Some(n.clone()),
-        Expr::Member { base, field, .. } => flatten_ident(base).map(|b| format!("{b}.{field}")),
-        Expr::Index { base, .. } => flatten_ident(base),
-        _ => None,
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use crate::pipeline::plc::parser;
-
-    const VULN: &str = r#"
-FUNCTION_BLOCK CommCtrl
-VAR_INPUT
-    cmdIndex : INT;
-END_VAR
-VAR
-    Password : STRING := 'admin123';
-    buffer : ARRAY[0..15] OF INT;
-    Safety_Enable : BOOL := TRUE;
-    divisor : INT;
-    result : INT;
-END_VAR
-    Safety_Enable := FALSE;
-    result := 100 / divisor;
-    buffer[cmdIndex] := 1;
-    Modbus_Connect(IP := '192.168.0.10', PORT := 502, AUTH := FALSE);
-    IF cmdIndex > 100 THEN
-        JMP fault;
-    END_IF;
-fault:
-    result := 0;
-END_FUNCTION_BLOCK
-"#;
-
-    fn rule_ids(src: &str) -> Vec<&'static str> {
-        parser::parse(src)
-            .iter()
-            .flat_map(analyze)
-            .map(|h| h.rule_id)
-            .collect()
-    }
-
-    #[test]
-    fn vulnerable_program_triggers_every_rule() {
-        let ids = rule_ids(VULN);
-        for expected in [
-            "plc-hardcoded-credential",
-            "plc-default-password",
-            "plc-safety-bypass",
-            "plc-division-by-zero",
-            "plc-array-unchecked-index",
-            "plc-insecure-comm",
-            "plc-insecure-protocol-port",
-            "plc-unstructured-jump",
-        ] {
-            assert!(
-                ids.contains(&expected),
-                "expected rule {expected}, got {ids:?}"
-            );
-        }
-    }
-
-    #[test]
-    fn clean_program_has_no_findings() {
-        let clean = r#"
-PROGRAM Clean
-VAR
-    a : INT := 5;
-    b : INT := 3;
-    total : INT;
-END_VAR
-    IF b <> 0 THEN
-        total := a / b;
-    END_IF;
-END_PROGRAM
-"#;
-        assert!(rule_ids(clean).is_empty(), "clean program should be quiet");
-    }
-}
@@ -1,253 +0,0 @@
-//! Control-application dependency SBOM from a CODESYS `.projectarchive`.
-//!
-//! A `.projectarchive` is a ZIP that bundles the project plus its referenced
-//! libraries and the target runtime. Each referenced library is an entry whose
-//! path segment follows the CODESYS convention
-//! `Name, Major.Minor.Patch.Build (Company)` (e.g. `Standard, 3.5.18.0 (System)`,
-//! `CSV Utility SL, 1.9.0.0 (CODESYS)`); the runtime appears as a device-descriptor
-//! entry `CODESYS Control … <version> …`. We enumerate those entries — no binary
-//! parsing — and emit SBOM components tagged `pkg:codesys/…`, so the CVE pipeline
-//! can match them (the runtime `Cmp*` / `3SLicense` components carry real CODESYS
-//! CVEs).
-
-use std::collections::BTreeSet;
-use std::path::{Path, PathBuf};
-
-use compliance_core::models::SbomEntry;
-
-/// Collect the control-application SBOM from every `.projectarchive` reachable for
-/// a target: the ingested artifact file itself (an uploaded archive), plus any
-/// `*.projectarchive` committed inside the working tree — e.g. a git repo or an
-/// extracted source archive that ships the archive alongside its PLCopen XML / ST
-/// exports. Deduplicated by (name, version).
-pub fn collect_sbom(artifact_file: &Path, working_path: &Path, repo_id: &str) -> Vec<SbomEntry> {
-    let mut archives: Vec<PathBuf> = Vec::new();
-    if artifact_file.is_file() {
-        archives.push(artifact_file.to_path_buf());
-    }
-    for entry in walkdir::WalkDir::new(working_path)
-        .max_depth(8)
-        .into_iter()
-        .filter_map(|e| e.ok())
-    {
-        let p = entry.path();
-        if entry.file_type().is_file()
-            && p.extension()
-                .and_then(|x| x.to_str())
-                .is_some_and(|x| x.eq_ignore_ascii_case("projectarchive"))
-        {
-            archives.push(p.to_path_buf());
-        }
-    }
-    let mut seen: BTreeSet<(String, String)> = BTreeSet::new();
-    let mut out = Vec::new();
-    for a in archives {
-        for e in projectarchive_sbom(&a, repo_id) {
-            if seen.insert((e.name.clone(), e.version.clone())) {
-                out.push(e);
-            }
-        }
-    }
-    out
-}
-
-/// Extract CODESYS library + runtime components from a `.projectarchive` (a zip).
-/// Best-effort: returns empty if the file is not a readable zip (e.g. a bare
-/// `.st`/`.xml` project, which carries no library manifest).
-pub fn projectarchive_sbom(archive: &Path, repo_id: &str) -> Vec<SbomEntry> {
-    let Ok(file) = std::fs::File::open(archive) else {
-        return Vec::new();
-    };
-    let Ok(mut zip) = zip::ZipArchive::new(file) else {
-        return Vec::new();
-    };
-    let mut seen: BTreeSet<(String, String)> = BTreeSet::new();
-    let mut entries = Vec::new();
-    for i in 0..zip.len() {
-        let Ok(entry) = zip.by_index(i) else {
-            continue;
-        };
-        // Entry paths use `\` (Windows-authored) and/or `/` separators; the
-        // component id is one path segment.
-        for seg in entry.name().split(['/', '\\']) {
-            if let Some((name, version)) = parse_library(seg).or_else(|| parse_runtime(seg)) {
-                if seen.insert((name.clone(), version.clone())) {
-                    let mut e = SbomEntry::new(
-                        repo_id.to_string(),
-                        name.clone(),
-                        version.clone(),
-                        "codesys".to_string(),
-                    );
-                    e.purl = Some(format!(
-                        "pkg:codesys/{}@{version}",
-                        name.replace(' ', "%20")
-                    ));
-                    entries.push(e);
-                }
-            }
-        }
-    }
-    entries
-}
-
-/// `Name, X.Y.Z.W (Company)` → (name, version).
-fn parse_library(seg: &str) -> Option<(String, String)> {
-    let seg = seg.trim();
-    // Company is the trailing "(…)".
-    let open = seg.rfind(" (")?;
-    let rest = &seg[open + 2..];
-    let close = rest.find(')')?;
-    if rest[..close].trim().is_empty() {
-        return None;
-    }
-    let head = seg[..open].trim(); // "Name, X.Y.Z.W"
-    let comma = head.rfind(", ")?;
-    let name = head[..comma].trim().to_string();
-    let version = head[comma + 2..].trim().to_string();
-    if name.is_empty() || !is_dotted_version(&version) {
-        return None;
-    }
-    Some((name, version))
-}
-
-/// Device-descriptor entry `CODESYS Control … X.Y.Z.W …` → (runtime name, version).
-fn parse_runtime(seg: &str) -> Option<(String, String)> {
-    let seg = seg.trim();
-    if !seg.starts_with("CODESYS Control") {
-        return None;
-    }
-    let version = seg
-        .split_whitespace()
-        .find(|t| is_dotted_version(t))?
-        .to_string();
-    // The runtime name is the first field, before the run of padding spaces that
-    // precede the descriptor's numeric columns.
-    let name = seg.split("  ").next().unwrap_or(seg).trim().to_string();
-    if name.is_empty() {
-        return None;
-    }
-    Some((name, version))
-}
-
-/// A dotted numeric version with at least 3 components (`3.5.18.0`, `4.17.0.0`).
-fn is_dotted_version(s: &str) -> bool {
-    let parts: Vec<&str> = s.split('.').collect();
-    parts.len() >= 3
-        && parts
-            .iter()
-            .all(|p| !p.is_empty() && p.chars().all(|c| c.is_ascii_digit()))
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use std::collections::HashMap;
-    use std::io::Write;
-
-    /// Build a synthetic `.projectarchive` (zip) mirroring the real CODESYS entry
-    /// naming (verified against Proemion/codesys-examples): a native `.project`,
-    /// referenced libraries as `Name, Version (Company)` segments, and a runtime
-    /// device descriptor.
-    fn synthetic_archive(dir: &Path) -> std::path::PathBuf {
-        let path = dir.join("App.projectarchive");
-        write_synthetic_archive(&path);
-        path
-    }
-
-    fn write_synthetic_archive(path: &Path) {
-        let file = std::fs::File::create(path).expect("create");
-        let mut zip = zip::ZipWriter::new(file);
-        let opts: zip::write::SimpleFileOptions = Default::default();
-        let names = [
-            "App.project",
-            r"{b0b5}\App.Device.Plc.compileinfo",
-            r"{e179}\Standard, 3.5.18.0 (System)   standard.compiled-library-v3",
-            r"{e179}\Util, 3.5.21.0 (System)   util.compiled-library-v3",
-            r"{e179}\CSV Utility SL, 1.9.0.0 (CODESYS)   csv utility sl.compiled-library-v3",
-            r"{e179}\3SLicense, 3.5.20.0 (CODESYS)   3slicense.compiled-library-v3",
-            r"{0c63}\CODESYS Control for Linux ARM SL     0000 0006     4.17.0.0     4096     .zip",
-        ];
-        for n in names {
-            zip.start_file(n, opts).expect("start");
-            zip.write_all(b"x").expect("write");
-        }
-        zip.finish().expect("finish");
-    }
-
-    #[test]
-    fn extracts_libraries_and_runtime_from_projectarchive() {
-        let tmp = std::env::temp_dir().join(format!("cs-plc-sbom-{}", uuid::Uuid::new_v4()));
-        std::fs::create_dir_all(&tmp).expect("mkdir");
-        let archive = synthetic_archive(&tmp);
-
-        let entries = projectarchive_sbom(&archive, "plc-target");
-        let by_name: HashMap<&str, &SbomEntry> =
-            entries.iter().map(|e| (e.name.as_str(), e)).collect();
-
-        // Libraries with their versions.
-        assert_eq!(
-            by_name.get("Standard").map(|e| e.version.as_str()),
-            Some("3.5.18.0")
-        );
-        assert_eq!(
-            by_name.get("Util").map(|e| e.version.as_str()),
-            Some("3.5.21.0")
-        );
-        assert_eq!(
-            by_name.get("CSV Utility SL").map(|e| e.version.as_str()),
-            Some("1.9.0.0"),
-            "multi-word library names must parse"
-        );
-        assert!(by_name.contains_key("3SLicense"));
-
-        // The runtime, from the device descriptor.
-        assert_eq!(
-            by_name
-                .get("CODESYS Control for Linux ARM SL")
-                .map(|e| e.version.as_str()),
-            Some("4.17.0.0")
-        );
-
-        // Every component is CODESYS-tagged with a purl the CVE pipeline can match,
-        // and the native `.project` / compileinfo are not mistaken for components.
-        for e in &entries {
-            assert_eq!(e.package_manager, "codesys");
-            assert!(e.purl.as_deref().unwrap_or("").starts_with("pkg:codesys/"));
-        }
-        assert!(!by_name.contains_key("App"));
-
-        let _ = std::fs::remove_dir_all(&tmp);
-    }
-
-    #[test]
-    fn collect_sbom_finds_a_projectarchive_committed_in_a_git_tree() {
-        let tmp = std::env::temp_dir().join(format!("cs-plc-collect-{}", uuid::Uuid::new_v4()));
-        let src = tmp.join("clone/src");
-        std::fs::create_dir_all(&src).expect("mkdir");
-        // Simulate a git clone that commits the archive alongside its exports.
-        write_synthetic_archive(&src.join("PumpStation.projectarchive"));
-
-        // The artifact "file" is a git URL (not a real file), so the SBOM must
-        // come from walking the cloned tree.
-        let entries = collect_sbom(Path::new("https://git.example/plc.git"), &tmp, "t");
-        let names: std::collections::HashSet<&str> =
-            entries.iter().map(|e| e.name.as_str()).collect();
-        assert!(
-            names.contains("Standard"),
-            "found libs in the committed archive"
-        );
-        assert!(names.contains("CODESYS Control for Linux ARM SL"));
-
-        let _ = std::fs::remove_dir_all(&tmp);
-    }
-
-    #[test]
-    fn non_zip_file_yields_no_sbom() {
-        let tmp = std::env::temp_dir().join(format!("cs-plc-sbom-st-{}", uuid::Uuid::new_v4()));
-        std::fs::create_dir_all(&tmp).expect("mkdir");
-        let st = tmp.join("prog.st");
-        std::fs::write(&st, "PROGRAM P\nVAR x : INT; END_VAR\nEND_PROGRAM\n").expect("write");
-        assert!(projectarchive_sbom(&st, "t").is_empty());
-        let _ = std::fs::remove_dir_all(&tmp);
-    }
-}
@@ -1,7 +1,5 @@
-use crate::pipeline::repo_view::RepoView;
 use compliance_core::models::*;

-use super::dedup::compute_fingerprint;
 use super::orchestrator::PipelineOrchestrator;
 use crate::error::AgentError;
 use crate::pipeline::code_review::CodeReviewScanner;
@@ -15,7 +13,7 @@ impl PipelineOrchestrator {
    #[tracing::instrument(skip_all, fields(repo_id = %repo_id, pr_number))]
    pub async fn run_pr_review(
        &self,
-        repo: &RepoView,
+        repo: &TrackedRepository,
        repo_id: &str,
        pr_number: u64,
        base_sha: &str,
@@ -91,37 +89,12 @@ impl PipelineOrchestrator {
            return Ok(());
        }

-        // Dedup findings by fingerprint to avoid duplicate comments
-        let mut seen_fps = std::collections::HashSet::new();
-        let mut unique_findings: Vec<&Finding> = Vec::new();
-        for finding in &pr_findings {
-            let fp = compute_fingerprint(&[
-                repo_id,
-                &pr_number.to_string(),
-                finding.file_path.as_deref().unwrap_or(""),
-                &finding.line_number.unwrap_or(0).to_string(),
-                &finding.title,
-            ]);
-            if seen_fps.insert(fp) {
-                unique_findings.push(finding);
-            }
-        }
-
-        let pr_findings = unique_findings;
-
        // Build review comments from findings
        let mut review_comments = Vec::new();
        for finding in &pr_findings {
            if let (Some(path), Some(line)) = (&finding.file_path, finding.line_number) {
-                let fp = compute_fingerprint(&[
-                    repo_id,
-                    &pr_number.to_string(),
-                    path,
-                    &line.to_string(),
-                    &finding.title,
-                ]);
                let comment_body = format!(
-                    "**[{}] {}**\n\n{}\n\n*Scanner: {} | {}*\n\n<!-- compliance-fp:{fp} -->",
+                    "**[{}] {}**\n\n{}\n\n*Scanner: {} | {}*",
                    finding.severity,
                    finding.title,
                    finding.description,
@@ -150,17 +123,6 @@ impl PipelineOrchestrator {
                .join("\n"),
        );

-        if review_comments.is_empty() {
-            // All findings were on files/lines we can't comment on inline
-            if let Err(e) = tracker
-                .create_pr_review(owner, tracker_repo_name, pr_number, &summary, Vec::new())
-                .await
-            {
-                tracing::warn!("[{repo_id}] Failed to post PR review summary: {e}");
-            }
-            return Ok(());
-        }
-
        if let Err(e) = tracker
            .create_pr_review(
                owner,
@@ -1,74 +0,0 @@
-//! `RepoView` — an internal, non-persisted view of a code target for the scan
-//! pipeline.
-//!
-//! It replaces the old persisted `TrackedRepository` model. The pipeline
-//! (SAST → SBOM → CVE → triage → issues → DAST, and PR review) only ever needs a
-//! flat bundle of git + issue-tracker + auth fields; those are projected from an
-//! [`OnboardedTarget`] and its code [`Artifact`] by [`RepoView::from_target`].
-//! Nothing here is written to Mongo — onboarded targets are the sole persisted
-//! entity.
-
-use compliance_core::models::{Artifact, OnboardedTarget, TrackerType};
-
-/// A flat, pipeline-facing view of a code target. Built from an onboarded
-/// target; never persisted.
-#[derive(Debug, Clone)]
-pub struct RepoView {
-    /// The onboarded target's id (used as `repo_id` across findings/sbom/etc.).
-    pub id: Option<mongodb::bson::oid::ObjectId>,
-    pub name: String,
-    pub git_url: String,
-    pub default_branch: String,
-    pub local_path: Option<String>,
-    pub scan_schedule: Option<String>,
-    pub webhook_enabled: bool,
-    pub webhook_secret: Option<String>,
-    pub tracker_type: Option<TrackerType>,
-    pub tracker_owner: Option<String>,
-    pub tracker_repo: Option<String>,
-    pub tracker_token: Option<String>,
-    pub auth_token: Option<String>,
-    pub auth_username: Option<String>,
-    pub last_scanned_commit: Option<String>,
-    pub findings_count: u32,
-}
-
-impl RepoView {
-    /// Project an onboarded target + its code artifact into a pipeline view.
-    pub fn from_target(target: &OnboardedTarget, code: &Artifact) -> Self {
-        let mut view = Self {
-            id: target.id,
-            name: target.name.clone(),
-            git_url: code.source_ref.clone(),
-            default_branch: "main".to_string(),
-            local_path: None,
-            scan_schedule: target.scan_schedule.clone(),
-            webhook_enabled: target.webhook_enabled,
-            webhook_secret: target.webhook_secret.clone(),
-            tracker_type: None,
-            tracker_owner: None,
-            tracker_repo: None,
-            tracker_token: None,
-            auth_token: None,
-            auth_username: None,
-            last_scanned_commit: None,
-            findings_count: target.findings_count,
-        };
-        if let Some(git) = &code.git {
-            view.default_branch = git.default_branch.clone();
-            view.last_scanned_commit = git.last_scanned_commit.clone();
-            view.local_path = git.local_path.clone();
-        }
-        if let Some(auth) = &code.auth {
-            view.auth_token = auth.secret.clone();
-            view.auth_username = auth.username.clone();
-        }
-        if let Some(it) = &target.scan_config.issue_tracker {
-            view.tracker_type = it.tracker_type.clone();
-            view.tracker_owner = it.owner.clone();
-            view.tracker_repo = it.repo.clone();
-            view.tracker_token = it.token.clone();
-        }
-        view
-    }
-}
@@ -5,26 +5,20 @@ use compliance_core::CoreError;

 #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
 pub(super) async fn run_syft(repo_path: &Path, repo_id: &str) -> Result<Vec<SbomEntry>, CoreError> {
-    let output = tokio::time::timeout(
-        std::time::Duration::from_secs(300),
-        tokio::process::Command::new("syft")
-            .arg(repo_path)
-            .args(["-o", "cyclonedx-json"])
-            .env("SYFT_GOLANG_SEARCH_REMOTE_LICENSES", "true")
-            .env("SYFT_JAVASCRIPT_SEARCH_REMOTE_LICENSES", "true")
-            .env("SYFT_PYTHON_SEARCH_REMOTE_LICENSES", "true")
-            .env("SYFT_JAVA_USE_NETWORK", "true")
-            .output(),
-    )
-    .await
-    .map_err(|_| CoreError::Scanner {
-        scanner: "syft".to_string(),
-        source: "timed out after 5 minutes".into(),
-    })?
-    .map_err(|e| CoreError::Scanner {
-        scanner: "syft".to_string(),
-        source: Box::new(e),
-    })?;
+    let output = tokio::process::Command::new("syft")
+        .arg(repo_path)
+        .args(["-o", "cyclonedx-json"])
+        // Enable remote license lookups for all ecosystems
+        .env("SYFT_GOLANG_SEARCH_REMOTE_LICENSES", "true")
+        .env("SYFT_JAVASCRIPT_SEARCH_REMOTE_LICENSES", "true")
+        .env("SYFT_PYTHON_SEARCH_REMOTE_LICENSES", "true")
+        .env("SYFT_JAVA_USE_NETWORK", "true")
+        .output()
+        .await
+        .map_err(|e| CoreError::Scanner {
+            scanner: "syft".to_string(),
+            source: Box::new(e),
+        })?;

    if !output.status.success() {
        let stderr = String::from_utf8_lossy(&output.stderr);
@@ -1,4 +1,4 @@
-use std::path::{Path, PathBuf};
+use std::path::Path;

 use compliance_core::models::{Finding, ScanType, Severity};
 use compliance_core::traits::{ScanOutput, Scanner};
@@ -6,30 +6,6 @@ use compliance_core::CoreError;

 use crate::pipeline::dedup;

-/// Custom CRA-control detectors bundled into the binary and staged to a temp file
-/// at scan time so semgrep can `--config` them alongside the auto ruleset. These
-/// cover controls no off-the-shelf rule digs out (secure defaults, weak password
-/// hashing, insecure session cookies, weak data-at-rest ciphers); each rule id is
-/// keyed back to its control by the `control-map` LUT.
-const CRA_RULES: &str = include_str!("../../rules/cra_semgrep.yaml");
-
-/// Write the bundled CRA rules to a stable temp path (atomic: unique tmp +
-/// rename). Returns `None` on failure — the scan then runs with auto rules only.
-async fn stage_cra_rules() -> Option<PathBuf> {
-    let dir = std::env::temp_dir();
-    let path = dir.join("compliance-cra-semgrep.yaml");
-    let tmp = dir.join(format!("compliance-cra-semgrep.{}.tmp", std::process::id()));
-    if let Err(e) = tokio::fs::write(&tmp, CRA_RULES).await {
-        tracing::warn!(error = %e, "failed to stage custom CRA semgrep rules; using auto rules only");
-        return None;
-    }
-    if let Err(e) = tokio::fs::rename(&tmp, &path).await {
-        tracing::warn!(error = %e, "failed to stage custom CRA semgrep rules; using auto rules only");
-        return None;
-    }
-    Some(path)
-}
-
 pub struct SemgrepScanner;

 impl Scanner for SemgrepScanner {
@@ -43,22 +19,11 @@ impl Scanner for SemgrepScanner {

    #[tracing::instrument(skip_all)]
    async fn scan(&self, repo_path: &Path, repo_id: &str) -> Result<ScanOutput, CoreError> {
-        let cra_rules = stage_cra_rules().await;
-        let mut command = tokio::process::Command::new("semgrep");
-        command.arg("--config=auto");
-        if let Some(path) = &cra_rules {
-            command.arg(format!("--config={}", path.display()));
-        }
-        command
-            .args(["--json", "--quiet", "--max-memory", "500", "--jobs", "1"])
-            .arg(repo_path);
-
-        let output = tokio::time::timeout(std::time::Duration::from_secs(600), command.output())
+        let output = tokio::process::Command::new("semgrep")
+            .args(["--config=auto", "--json", "--quiet"])
+            .arg(repo_path)
+            .output()
            .await
-            .map_err(|_| CoreError::Scanner {
-                scanner: "semgrep".to_string(),
-                source: "timed out after 10 minutes".into(),
-            })?
            .map_err(|e| CoreError::Scanner {
                scanner: "semgrep".to_string(),
                source: Box::new(e),
@@ -102,7 +67,10 @@ impl Scanner for SemgrepScanner {
                finding.file_path = Some(r.path);
                finding.line_number = Some(r.start.line);
                finding.code_snippet = Some(r.extra.lines);
-                finding.cwe = r.extra.metadata.as_ref().and_then(extract_cwe);
+                finding.cwe = r
+                    .extra
+                    .metadata
+                    .and_then(|m| m.get("cwe").and_then(|v| v.as_str()).map(|s| s.to_string()));
                finding
            })
            .collect();
@@ -141,34 +109,10 @@ struct SemgrepExtra {
    metadata: Option<serde_json::Value>,
 }

-/// semgrep emits `metadata.cwe` as a list of strings like
-/// `"CWE-798: Use of Hard-coded Credentials"` (occasionally a bare string). Take
-/// the first entry and normalise it to just the `CWE-NNN` id.
-fn extract_cwe(metadata: &serde_json::Value) -> Option<String> {
-    let raw = metadata.get("cwe")?;
-    let text = match raw {
-        serde_json::Value::Array(items) => items.first()?.as_str()?,
-        serde_json::Value::String(s) => s.as_str(),
-        _ => return None,
-    };
-    let id = text.split(':').next().unwrap_or(text).trim();
-    (!id.is_empty()).then(|| id.to_string())
-}
-
 #[cfg(test)]
 mod tests {
    use super::*;

-    #[test]
-    fn extract_cwe_handles_list_and_normalises() {
-        let md = serde_json::json!({"cwe": ["CWE-798: Use of Hard-coded Credentials"]});
-        assert_eq!(extract_cwe(&md).as_deref(), Some("CWE-798"));
-        let bare = serde_json::json!({"cwe": "CWE-89"});
-        assert_eq!(extract_cwe(&bare).as_deref(), Some("CWE-89"));
-        let none = serde_json::json!({"severity": "ERROR"});
-        assert_eq!(extract_cwe(&none), None);
-    }
-
    #[test]
    fn deserialize_semgrep_output() {
        let json = r#"{
@@ -6,16 +6,11 @@ use compliance_core::models::embedding::{CodeEmbedding, EmbeddingBuildRun, Embed
 use compliance_core::models::graph::CodeNode;
 use compliance_graph::graph::chunking::extract_chunks;
 use compliance_graph::graph::embedding_store::EmbeddingStore;
-use futures_util::stream::{FuturesUnordered, StreamExt};
 use tracing::{error, info};

 use crate::error::AgentError;
 use crate::llm::LlmClient;

-const EMBED_BATCH_SIZE: usize = 20;
-const EMBED_CONCURRENCY: usize = 4;
-const EMBED_FLUSH_EVERY: usize = 200;
-
 /// RAG pipeline for building embeddings and performing retrieval
 pub struct RagPipeline {
    llm: Arc<LlmClient>,
@@ -82,33 +77,25 @@ impl RagPipeline {
            .await
            .map_err(|e| AgentError::Other(format!("Failed to delete old embeddings: {e}")))?;

-        // Step 3: Batch embed with bounded concurrency. Flush to Mongo and
-        // update progress periodically so the dashboard can show live status.
-        let mut pending = Vec::with_capacity(EMBED_FLUSH_EVERY);
+        // Step 3: Batch embed (small batches to stay within model limits)
+        let batch_size = 20;
+        let mut all_embeddings = Vec::new();
        let mut embedded_count = 0u32;

-        // Build the list of batch indices to process.
-        let batches: Vec<(usize, usize)> = (0..chunks.len())
-            .step_by(EMBED_BATCH_SIZE)
-            .map(|start| (start, (start + EMBED_BATCH_SIZE).min(chunks.len())))
-            .collect();
+        for batch_start in (0..chunks.len()).step_by(batch_size) {
+            let batch_end = (batch_start + batch_size).min(chunks.len());
+            let batch_chunks = &chunks[batch_start..batch_end];

-        let mut batch_iter = batches.into_iter();
-        let mut in_flight = FuturesUnordered::new();
+            // Prepare texts: context_header + content
+            let texts: Vec<String> = batch_chunks
+                .iter()
+                .map(|c| format!("{}\n{}", c.context_header, c.content))
+                .collect();

-        // Prime up to EMBED_CONCURRENCY batches.
-        for _ in 0..EMBED_CONCURRENCY {
-            if let Some((start, end)) = batch_iter.next() {
-                in_flight.push(self.embed_batch(&chunks[start..end], start, end));
-            }
-        }
-
-        while let Some(result) = in_flight.next().await {
-            match result {
-                Ok((start, end, vectors)) => {
-                    let batch_chunks = &chunks[start..end];
+            match self.llm.embed(texts).await {
+                Ok(vectors) => {
                    for (chunk, embedding) in batch_chunks.iter().zip(vectors) {
-                        pending.push(CodeEmbedding {
+                        all_embeddings.push(CodeEmbedding {
                            id: None,
                            repo_id: repo_id.to_string(),
                            graph_build_id: graph_build_id.to_string(),
@@ -126,45 +113,9 @@ impl RagPipeline {
                        });
                    }
                    embedded_count += batch_chunks.len() as u32;
-
-                    // Flush pending embeddings to Mongo periodically and update progress.
-                    if pending.len() >= EMBED_FLUSH_EVERY {
-                        self.embedding_store
-                            .store_embeddings(&pending)
-                            .await
-                            .map_err(|e| {
-                                AgentError::Other(format!("Failed to store embeddings: {e}"))
-                            })?;
-                        pending.clear();
-                    }
-
-                    // Always update the progress counter on the build doc — even if
-                    // we haven't flushed embeddings yet — so the UI shows movement.
-                    if let Err(e) = self
-                        .embedding_store
-                        .update_build(
-                            repo_id,
-                            graph_build_id,
-                            EmbeddingBuildStatus::Running,
-                            embedded_count,
-                            None,
-                        )
-                        .await
-                    {
-                        error!("[{repo_id}] Failed to update build progress: {e}");
-                    }
-
-                    // Queue the next batch to keep concurrency saturated.
-                    if let Some((s, e)) = batch_iter.next() {
-                        in_flight.push(self.embed_batch(&chunks[s..e], s, e));
-                    }
                }
                Err(e) => {
                    error!("[{repo_id}] Embedding batch failed: {e}");
-                    // Flush whatever we have so partial progress isn't lost.
-                    if !pending.is_empty() {
-                        let _ = self.embedding_store.store_embeddings(&pending).await;
-                    }
                    build.status = EmbeddingBuildStatus::Failed;
                    build.error_message = Some(e.to_string());
                    build.completed_at = Some(Utc::now());
@@ -183,13 +134,11 @@ impl RagPipeline {
            }
        }

-        // Step 4: Flush any remaining embeddings
-        if !pending.is_empty() {
-            self.embedding_store
-                .store_embeddings(&pending)
-                .await
-                .map_err(|e| AgentError::Other(format!("Failed to store embeddings: {e}")))?;
-        }
+        // Step 4: Store all embeddings
+        self.embedding_store
+            .store_embeddings(&all_embeddings)
+            .await
+            .map_err(|e| AgentError::Other(format!("Failed to store embeddings: {e}")))?;

        // Step 5: Update build status
        build.status = EmbeddingBuildStatus::Completed;
@@ -212,21 +161,4 @@ impl RagPipeline {
        );
        Ok(build)
    }
-
-    /// Embed one batch of chunks. Returns the (start, end, vectors) tuple so
-    /// out-of-order completion from `FuturesUnordered` can still be reconciled
-    /// against the original chunk slice.
-    async fn embed_batch(
-        &self,
-        batch_chunks: &[compliance_graph::graph::chunking::CodeChunk],
-        start: usize,
-        end: usize,
-    ) -> Result<(usize, usize, Vec<Vec<f64>>), AgentError> {
-        let texts: Vec<String> = batch_chunks
-            .iter()
-            .map(|c| format!("{}\n{}", c.context_header, c.content))
-            .collect();
-        let vectors = self.llm.embed(texts).await?;
-        Ok((start, end, vectors))
-    }
 }
@@ -4,21 +4,8 @@ use tokio_cron_scheduler::{Job, JobScheduler};
 use compliance_core::models::ScanTrigger;

 use crate::agent::ComplianceAgent;
-use crate::database::Database;
 use crate::error::AgentError;

-/// Default tenant the scheduler runs against when neither the tenant
-/// registry nor `SCHEDULER_TENANT_IDS` are configured. Matches the
-/// dev-injector default so a bare `cargo run` has the scheduler
-/// scanning whatever lives in `<prefix>_dev`.
-const DEFAULT_SCHEDULER_TENANT_ID: &str = "dev";
-
-/// Request timeout when fetching the live tenant list from the
-/// registry. Kept short — if the registry is slow we'd rather fall
-/// back to env-configured ids and finish the tick than block the
-/// scheduler loop.
-const REGISTRY_FETCH_TIMEOUT_SECS: u64 = 5;
-
 pub async fn start_scheduler(agent: &ComplianceAgent) -> Result<(), AgentError> {
    let sched = JobScheduler::new()
        .await
@@ -31,14 +18,7 @@ pub async fn start_scheduler(agent: &ComplianceAgent) -> Result<(), AgentError>
        let agent = scan_agent.clone();
        Box::pin(async move {
            tracing::info!("Scheduled scan triggered");
-            let tenants = scheduler_tenants(&agent).await;
-            tracing::debug!(
-                tenant_count = tenants.len(),
-                "Scheduled scan: tenants resolved"
-            );
-            for tenant_id in tenants {
-                scan_all_repos(&agent, &tenant_id).await;
-            }
+            scan_all_repos(&agent).await;
        })
    })
    .map_err(|e| AgentError::Scheduler(format!("Failed to create scan job: {e}")))?;
@@ -54,14 +34,7 @@ pub async fn start_scheduler(agent: &ComplianceAgent) -> Result<(), AgentError>
        let agent = cve_agent.clone();
        Box::pin(async move {
            tracing::info!("CVE monitor triggered");
-            let tenants = scheduler_tenants(&agent).await;
-            tracing::debug!(
-                tenant_count = tenants.len(),
-                "CVE monitor: tenants resolved"
-            );
-            for tenant_id in tenants {
-                monitor_cves(&agent, &tenant_id).await;
-            }
+            monitor_cves(&agent).await;
        })
    })
    .map_err(|e| AgentError::Scheduler(format!("Failed to create CVE monitor job: {e}")))?;
@@ -75,14 +48,8 @@ pub async fn start_scheduler(agent: &ComplianceAgent) -> Result<(), AgentError>
        .await
        .map_err(|e| AgentError::Scheduler(format!("Failed to start scheduler: {e}")))?;

-    let tenants = scheduler_tenants(agent).await;
-    let source = if agent.config.tenant_registry_url.is_some() {
-        "tenant-registry (env fallback)"
-    } else {
-        "env (SCHEDULER_TENANT_IDS)"
-    };
    tracing::info!(
-        "Scheduler started: scans='{}', CVE monitor='{}', tenant source={source}, tenants={tenants:?}",
+        "Scheduler started: scans='{}', CVE monitor='{}'",
        agent.config.scan_schedule,
        agent.config.cve_monitor_schedule,
    );
@@ -93,382 +60,46 @@ pub async fn start_scheduler(agent: &ComplianceAgent) -> Result<(), AgentError>
    }
 }

-/// Tenants the scheduler iterates each tick.
-///
-/// Resolution order:
-/// 1. **Tenant registry** at `agent.config.tenant_registry_url`
-///    (`GET /v1/tenants`). Fresh on every tick — picks up newly
-///    provisioned tenants without an agent restart.
-/// 2. **`SCHEDULER_TENANT_IDS`** env (comma-separated) — fallback when
-///    the registry is unreachable, the response is malformed, or no
-///    registry URL is configured.
-/// 3. **`DEFAULT_SCHEDULER_TENANT_ID`** (`"dev"`) — last-ditch fallback
-///    so the scheduler keeps doing something useful in dev.
-///
-/// We never panic out of this function — the scheduler must keep
-/// firing even if the registry is offline.
-async fn scheduler_tenants(agent: &ComplianceAgent) -> Vec<String> {
-    if let Some(url) = agent.config.tenant_registry_url.as_deref() {
-        match fetch_tenants_from_registry(&agent.http, url).await {
-            Ok(v) if !v.is_empty() => return v,
-            Ok(_) => {
-                tracing::warn!("tenant-registry returned empty list; falling back to env");
-            }
-            Err(e) => {
-                tracing::warn!(
-                    url = %url,
-                    error = %e,
-                    "tenant-registry fetch failed; falling back to env"
-                );
-            }
-        }
-    }
-    tenants_from_env()
-}
-
-fn tenants_from_env() -> Vec<String> {
-    std::env::var("SCHEDULER_TENANT_IDS")
-        .ok()
-        .map(|s| {
-            s.split(',')
-                .map(str::trim)
-                .filter(|s| !s.is_empty())
-                .map(String::from)
-                .collect::<Vec<_>>()
-        })
-        .filter(|v| !v.is_empty())
-        .unwrap_or_else(|| vec![DEFAULT_SCHEDULER_TENANT_ID.to_string()])
-}
-
-/// Shape we accept from the registry. Liberal in what we accept:
-/// the registry can return any field shape as long as either `id` or
-/// `tenant_id` is present. Other fields are ignored.
-#[derive(serde::Deserialize)]
-struct RegistryTenant {
-    #[serde(alias = "tenant_id")]
-    id: String,
-    /// Filter out non-running tenants if status is present. Missing
-    /// status defaults to "active" so older registry deployments keep
-    /// working.
-    #[serde(default = "default_status")]
-    status: String,
-}
-
-fn default_status() -> String {
-    "active".to_string()
-}
-
-#[derive(serde::Deserialize)]
-struct RegistryListResponse {
-    data: Vec<RegistryTenant>,
-}
-
-async fn fetch_tenants_from_registry(
-    http: &reqwest::Client,
-    base_url: &str,
-) -> Result<Vec<String>, String> {
-    let url = format!("{}/v1/tenants", base_url.trim_end_matches('/'));
-    let resp = http
-        .get(&url)
-        .timeout(std::time::Duration::from_secs(REGISTRY_FETCH_TIMEOUT_SECS))
-        .send()
-        .await
-        .map_err(|e| format!("request failed: {e}"))?;
-    if !resp.status().is_success() {
-        return Err(format!("registry returned {}", resp.status()));
-    }
-    let body: RegistryListResponse = resp
-        .json()
-        .await
-        .map_err(|e| format!("invalid JSON: {e}"))?;
-    Ok(filter_active(body.data))
-}
-
-/// Frozen/Archived tenants don't need scheduled scans; the M7.1
-/// status gate would 402/410 anyway. Skip them so we don't waste
-/// cycles. Active / trial / demo / anything-else-unknown all run.
-fn filter_active(rows: Vec<RegistryTenant>) -> Vec<String> {
-    rows.into_iter()
-        .filter(|t| !matches!(t.status.as_str(), "frozen" | "archived"))
-        .map(|t| t.id)
-        .collect()
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    fn tenant(id: &str, status: &str) -> RegistryTenant {
-        RegistryTenant {
-            id: id.to_string(),
-            status: status.to_string(),
-        }
-    }
-
-    #[test]
-    fn filter_active_keeps_running_skips_frozen_archived() {
-        let rows = vec![
-            tenant("a", "active"),
-            tenant("b", "trial"),
-            tenant("c", "demo"),
-            tenant("d", "frozen"),
-            tenant("e", "archived"),
-            tenant("f", "weird-but-not-known-dead"),
-        ];
-        let out = filter_active(rows);
-        assert_eq!(out, vec!["a", "b", "c", "f"]);
-    }
-
-    #[test]
-    fn deserialize_registry_response_accepts_id_or_tenant_id() {
-        let body = r#"{"data":[
-            {"id":"a","status":"active"},
-            {"tenant_id":"b","status":"trial"},
-            {"id":"c"}
-        ]}"#;
-        let parsed: RegistryListResponse = serde_json::from_str(body).unwrap();
-        assert_eq!(parsed.data.len(), 3);
-        assert_eq!(parsed.data[0].id, "a");
-        assert_eq!(parsed.data[1].id, "b");
-        assert_eq!(parsed.data[2].id, "c");
-        // Default status for the third entry should be "active"
-        assert_eq!(parsed.data[2].status, "active");
-    }
-
-    /// Combined into a single test: cargo runs tests in parallel and
-    /// env vars are process-global, so two separate tests touching
-    /// `SCHEDULER_TENANT_IDS` race each other. Doing both checks in
-    /// one test keeps them in a deterministic order.
-    #[test]
-    fn tenants_from_env_resolution() {
-        std::env::remove_var("SCHEDULER_TENANT_IDS");
-        assert_eq!(
-            tenants_from_env(),
-            vec![DEFAULT_SCHEDULER_TENANT_ID.to_string()],
-            "unset → default"
-        );
-
-        std::env::set_var("SCHEDULER_TENANT_IDS", "acme, globex ,,hello");
-        let out = tenants_from_env();
-        std::env::remove_var("SCHEDULER_TENANT_IDS");
-        assert_eq!(
-            out,
-            vec!["acme", "globex", "hello"],
-            "splits + trims + drops empty"
-        );
-
-        std::env::set_var("SCHEDULER_TENANT_IDS", "");
-        let out = tenants_from_env();
-        std::env::remove_var("SCHEDULER_TENANT_IDS");
-        assert_eq!(
-            out,
-            vec![DEFAULT_SCHEDULER_TENANT_ID.to_string()],
-            "empty → default"
-        );
-    }
-}
-
-/// Resolve the per-tenant database. Logs and returns `None` on failure
-/// so the loop in the caller can continue with other tenants.
-async fn tenant_db(agent: &ComplianceAgent, tenant_id: &str) -> Option<Database> {
-    match agent.db_pool.for_tenant_id(tenant_id).await {
-        Ok(db) => Some(db),
-        Err(e) => {
-            tracing::error!("Scheduler: cannot open tenant database '{tenant_id}': {e}");
-            None
-        }
-    }
-}
-
-async fn scan_all_repos(agent: &ComplianceAgent, tenant_id: &str) {
+async fn scan_all_repos(agent: &ComplianceAgent) {
    use futures_util::StreamExt;

-    let db = match tenant_db(agent, tenant_id).await {
-        Some(db) => db,
-        None => return,
-    };
-
-    let cursor = match db.onboarded_targets().find(doc! {}).await {
+    let cursor = match agent.db.repositories().find(doc! {}).await {
        Ok(c) => c,
        Err(e) => {
-            tracing::error!("Failed to list targets for tenant '{tenant_id}': {e}");
+            tracing::error!("Failed to list repos for scheduled scan: {e}");
            return;
        }
    };

-    let targets: Vec<_> = cursor.filter_map(|r| async { r.ok() }).collect().await;
+    let repos: Vec<_> = cursor.filter_map(|r| async { r.ok() }).collect().await;

-    for target in targets {
-        let target_id = target.id.map(|id| id.to_hex()).unwrap_or_default();
-        if let Err(e) = agent
-            .run_target_scan(tenant_id, &target_id, ScanTrigger::Scheduled)
-            .await
-        {
-            tracing::error!(
-                "Scheduled scan failed for {} (tenant '{tenant_id}'): {e}",
-                target.name
-            );
+    for repo in repos {
+        let repo_id = repo.id.map(|id| id.to_hex()).unwrap_or_default();
+        if let Err(e) = agent.run_scan(&repo_id, ScanTrigger::Scheduled).await {
+            tracing::error!("Scheduled scan failed for {}: {e}", repo.name);
        }
    }
 }

-async fn monitor_cves(agent: &ComplianceAgent, tenant_id: &str) {
-    use compliance_core::models::notification::{parse_severity, CveNotification};
-    use compliance_core::models::SbomEntry;
+async fn monitor_cves(agent: &ComplianceAgent) {
    use futures_util::StreamExt;

-    let db = match tenant_db(agent, tenant_id).await {
-        Some(db) => db,
-        None => return,
-    };
-
-    // Fetch all SBOM entries grouped by repo
-    let cursor = match db.sbom_entries().find(doc! {}).await {
+    // Re-scan all SBOM entries for new CVEs
+    let cursor = match agent.db.sbom_entries().find(doc! {}).await {
        Ok(c) => c,
        Err(e) => {
-            tracing::error!("CVE monitor: failed to list SBOM entries for '{tenant_id}': {e}");
+            tracing::error!("Failed to list SBOM entries for CVE monitoring: {e}");
            return;
        }
    };
-    let entries: Vec<SbomEntry> = cursor.filter_map(|r| async { r.ok() }).collect().await;
+
+    let entries: Vec<_> = cursor.filter_map(|r| async { r.ok() }).collect().await;
+
    if entries.is_empty() {
-        tracing::debug!("CVE monitor: no SBOM entries for tenant '{tenant_id}', skipping");
        return;
    }

-    tracing::info!(
-        "CVE monitor: checking {} dependencies for new CVEs (tenant '{tenant_id}')",
-        entries.len()
-    );
-
-    // Build a repo_id → repo_name lookup
-    let repo_ids: std::collections::HashSet<String> =
-        entries.iter().map(|e| e.repo_id.clone()).collect();
-    let mut repo_names: std::collections::HashMap<String, String> =
-        std::collections::HashMap::new();
-    for rid in &repo_ids {
-        if let Ok(oid) = mongodb::bson::oid::ObjectId::parse_str(rid) {
-            if let Ok(Some(repo)) = db.onboarded_targets().find_one(doc! { "_id": oid }).await {
-                repo_names.insert(rid.clone(), repo.name.clone());
-            }
-        }
-    }
-
-    // Use the existing CveScanner to query OSV.dev
-    let nvd_key = agent.config.nvd_api_key.as_ref().map(|k| {
-        use secrecy::ExposeSecret;
-        k.expose_secret().to_string()
-    });
-    let scanner = crate::pipeline::cve::CveScanner::new(
-        agent.http.clone(),
-        agent.config.searxng_url.clone(),
-        nvd_key,
-    );
-
-    // Group entries by repo for scanning
-    let mut entries_by_repo: std::collections::HashMap<String, Vec<SbomEntry>> =
-        std::collections::HashMap::new();
-    for entry in entries {
-        entries_by_repo
-            .entry(entry.repo_id.clone())
-            .or_default()
-            .push(entry);
-    }
-
-    let mut new_notifications = 0u32;
-
-    for (repo_id, mut repo_entries) in entries_by_repo {
-        let repo_name = repo_names
-            .get(&repo_id)
-            .cloned()
-            .unwrap_or_else(|| repo_id.clone());
-
-        // Scan dependencies for CVEs
-        let alerts = match scanner.scan_dependencies(&repo_id, &mut repo_entries).await {
-            Ok(a) => a,
-            Err(e) => {
-                tracing::warn!("CVE monitor: scan failed for {repo_name}: {e}");
-                continue;
-            }
-        };
-
-        // Upsert CVE alerts (existing logic)
-        for alert in &alerts {
-            let filter = doc! { "cve_id": &alert.cve_id, "repo_id": &alert.repo_id };
-            let update = doc! { "$setOnInsert": mongodb::bson::to_bson(alert).unwrap_or_default() };
-            let _ = db
-                .cve_alerts()
-                .update_one(filter, update)
-                .upsert(true)
-                .await;
-        }
-
-        // Update SBOM entries with discovered vulnerabilities
-        for entry in &repo_entries {
-            if entry.known_vulnerabilities.is_empty() {
-                continue;
-            }
-            if let Some(entry_id) = &entry.id {
-                let _ = db
-                    .sbom_entries()
-                    .update_one(
-                        doc! { "_id": entry_id },
-                        doc! { "$set": {
-                            "known_vulnerabilities": mongodb::bson::to_bson(&entry.known_vulnerabilities).unwrap_or_default(),
-                            "updated_at": mongodb::bson::DateTime::now(),
-                        }},
-                    )
-                    .await;
-            }
-        }
-
-        // Create notifications for NEW CVEs (dedup against existing notifications)
-        for alert in &alerts {
-            let filter = doc! {
-                "cve_id": &alert.cve_id,
-                "repo_id": &alert.repo_id,
-                "package_name": &alert.affected_package,
-                "package_version": &alert.affected_version,
-            };
-            // Only insert if not already exists (upsert with $setOnInsert)
-            let severity = parse_severity(alert.severity.as_deref(), alert.cvss_score);
-            let mut notification = CveNotification::new(
-                alert.cve_id.clone(),
-                repo_id.clone(),
-                repo_name.clone(),
-                alert.affected_package.clone(),
-                alert.affected_version.clone(),
-                severity,
-            );
-            notification.cvss_score = alert.cvss_score;
-            notification.summary = alert.summary.clone();
-            notification.url = Some(format!("https://osv.dev/vulnerability/{}", alert.cve_id));
-
-            let update = doc! {
-                "$setOnInsert": mongodb::bson::to_bson(&notification).unwrap_or_default()
-            };
-            match db
-                .cve_notifications()
-                .update_one(filter, update)
-                .upsert(true)
-                .await
-            {
-                Ok(result) if result.upserted_id.is_some() => {
-                    new_notifications += 1;
-                }
-                Err(e) => {
-                    tracing::warn!("CVE monitor: failed to create notification: {e}");
-                }
-                _ => {} // Already exists
-            }
-        }
-    }
-
-    if new_notifications > 0 {
-        tracing::info!(
-            "CVE monitor: created {new_notifications} new notification(s) for tenant '{tenant_id}'"
-        );
-    } else {
-        tracing::info!("CVE monitor: no new CVEs found for tenant '{tenant_id}'");
-    }
+    tracing::info!("CVE monitor: checking {} dependencies", entries.len());
+    // The actual CVE checking is handled by the CveScanner in the pipeline
+    // This is a simplified version that just logs the activity
 }
@@ -14,30 +14,24 @@ type HmacSha256 = Hmac<Sha256>;

 pub async fn handle_gitea_webhook(
    Extension(agent): Extension<Arc<ComplianceAgent>>,
-    Path((tenant_id, repo_id)): Path<(String, String)>,
+    Path(repo_id): Path<String>,
    headers: HeaderMap,
    body: Bytes,
 ) -> StatusCode {
-    // Look up the repo in the tenant's database to get its webhook secret
+    // Look up the repo to get its webhook secret
    let oid = match mongodb::bson::oid::ObjectId::parse_str(&repo_id) {
        Ok(oid) => oid,
        Err(_) => return StatusCode::NOT_FOUND,
    };
-    let db = match agent.db_pool.for_tenant_id(&tenant_id).await {
-        Ok(db) => db,
-        Err(e) => {
-            tracing::warn!("Gitea webhook: cannot open tenant database '{tenant_id}': {e}");
-            return StatusCode::NOT_FOUND;
-        }
-    };
-    let repo = match db
-        .onboarded_targets()
+    let repo = match agent
+        .db
+        .repositories()
        .find_one(mongodb::bson::doc! { "_id": oid })
        .await
    {
        Ok(Some(repo)) => repo,
        _ => {
-            tracing::warn!("Gitea webhook: repo {repo_id} not found in tenant '{tenant_id}'");
+            tracing::warn!("Gitea webhook: repo {repo_id} not found");
            return StatusCode::NOT_FOUND;
        }
    };
@@ -72,21 +66,15 @@ pub async fn handle_gitea_webhook(
        "push" => {
            let agent_clone = (*agent).clone();
            let repo_id = repo_id.clone();
-            let tenant_id = tenant_id.clone();
            tokio::spawn(async move {
-                tracing::info!(
-                    "Gitea push webhook: triggering scan for {repo_id} in tenant {tenant_id}"
-                );
-                if let Err(e) = agent_clone
-                    .run_scan(&tenant_id, &repo_id, ScanTrigger::Webhook)
-                    .await
-                {
+                tracing::info!("Gitea push webhook: triggering scan for {repo_id}");
+                if let Err(e) = agent_clone.run_scan(&repo_id, ScanTrigger::Webhook).await {
                    tracing::error!("Webhook-triggered scan failed: {e}");
                }
            });
            StatusCode::OK
        }
-        "pull_request" => handle_pull_request(agent, &tenant_id, &repo_id, &payload).await,
+        "pull_request" => handle_pull_request(agent, &repo_id, &payload).await,
        _ => {
            tracing::debug!("Gitea webhook: ignoring event '{event}'");
            StatusCode::OK
@@ -96,7 +84,6 @@ pub async fn handle_gitea_webhook(

 async fn handle_pull_request(
    agent: Arc<ComplianceAgent>,
-    tenant_id: &str,
    repo_id: &str,
    payload: &serde_json::Value,
 ) -> StatusCode {
@@ -119,14 +106,13 @@ async fn handle_pull_request(
    }

    let repo_id = repo_id.to_string();
-    let tenant_id = tenant_id.to_string();
    let head_sha = head_sha.to_string();
    let base_sha = base_sha.to_string();
    let agent_clone = (*agent).clone();
    tokio::spawn(async move {
        tracing::info!("Gitea PR webhook: reviewing PR #{pr_number} on {repo_id}");
        if let Err(e) = agent_clone
-            .run_pr_review(&tenant_id, &repo_id, pr_number, &base_sha, &head_sha)
+            .run_pr_review(&repo_id, pr_number, &base_sha, &head_sha)
            .await
        {
            tracing::error!("PR review failed for #{pr_number}: {e}");
@@ -14,30 +14,24 @@ type HmacSha256 = Hmac<Sha256>;

 pub async fn handle_github_webhook(
    Extension(agent): Extension<Arc<ComplianceAgent>>,
-    Path((tenant_id, repo_id)): Path<(String, String)>,
+    Path(repo_id): Path<String>,
    headers: HeaderMap,
    body: Bytes,
 ) -> StatusCode {
-    // Look up the repo in the tenant's database to get its webhook secret
+    // Look up the repo to get its webhook secret
    let oid = match mongodb::bson::oid::ObjectId::parse_str(&repo_id) {
        Ok(oid) => oid,
        Err(_) => return StatusCode::NOT_FOUND,
    };
-    let db = match agent.db_pool.for_tenant_id(&tenant_id).await {
-        Ok(db) => db,
-        Err(e) => {
-            tracing::warn!("GitHub webhook: cannot open tenant database '{tenant_id}': {e}");
-            return StatusCode::NOT_FOUND;
-        }
-    };
-    let repo = match db
-        .onboarded_targets()
+    let repo = match agent
+        .db
+        .repositories()
        .find_one(mongodb::bson::doc! { "_id": oid })
        .await
    {
        Ok(Some(repo)) => repo,
        _ => {
-            tracing::warn!("GitHub webhook: repo {repo_id} not found in tenant '{tenant_id}'");
+            tracing::warn!("GitHub webhook: repo {repo_id} not found");
            return StatusCode::NOT_FOUND;
        }
    };
@@ -72,21 +66,15 @@ pub async fn handle_github_webhook(
        "push" => {
            let agent_clone = (*agent).clone();
            let repo_id = repo_id.clone();
-            let tenant_id = tenant_id.clone();
            tokio::spawn(async move {
-                tracing::info!(
-                    "GitHub push webhook: triggering scan for {repo_id} in tenant {tenant_id}"
-                );
-                if let Err(e) = agent_clone
-                    .run_scan(&tenant_id, &repo_id, ScanTrigger::Webhook)
-                    .await
-                {
+                tracing::info!("GitHub push webhook: triggering scan for {repo_id}");
+                if let Err(e) = agent_clone.run_scan(&repo_id, ScanTrigger::Webhook).await {
                    tracing::error!("Webhook-triggered scan failed: {e}");
                }
            });
            StatusCode::OK
        }
-        "pull_request" => handle_pull_request(agent, &tenant_id, &repo_id, &payload).await,
+        "pull_request" => handle_pull_request(agent, &repo_id, &payload).await,
        _ => {
            tracing::debug!("GitHub webhook: ignoring event '{event}'");
            StatusCode::OK
@@ -96,7 +84,6 @@ pub async fn handle_github_webhook(

 async fn handle_pull_request(
    agent: Arc<ComplianceAgent>,
-    tenant_id: &str,
    repo_id: &str,
    payload: &serde_json::Value,
 ) -> StatusCode {
@@ -118,14 +105,13 @@ async fn handle_pull_request(
    }

    let repo_id = repo_id.to_string();
-    let tenant_id = tenant_id.to_string();
    let head_sha = head_sha.to_string();
    let base_sha = base_sha.to_string();
    let agent_clone = (*agent).clone();
    tokio::spawn(async move {
        tracing::info!("GitHub PR webhook: reviewing PR #{pr_number} on {repo_id}");
        if let Err(e) = agent_clone
-            .run_pr_review(&tenant_id, &repo_id, pr_number, &base_sha, &head_sha)
+            .run_pr_review(&repo_id, pr_number, &base_sha, &head_sha)
            .await
        {
            tracing::error!("PR review failed for #{pr_number}: {e}");
@@ -10,30 +10,24 @@ use crate::agent::ComplianceAgent;

 pub async fn handle_gitlab_webhook(
    Extension(agent): Extension<Arc<ComplianceAgent>>,
-    Path((tenant_id, repo_id)): Path<(String, String)>,
+    Path(repo_id): Path<String>,
    headers: HeaderMap,
    body: Bytes,
 ) -> StatusCode {
-    // Look up the repo in the tenant's database to get its webhook secret
+    // Look up the repo to get its webhook secret
    let oid = match mongodb::bson::oid::ObjectId::parse_str(&repo_id) {
        Ok(oid) => oid,
        Err(_) => return StatusCode::NOT_FOUND,
    };
-    let db = match agent.db_pool.for_tenant_id(&tenant_id).await {
-        Ok(db) => db,
-        Err(e) => {
-            tracing::warn!("GitLab webhook: cannot open tenant database '{tenant_id}': {e}");
-            return StatusCode::NOT_FOUND;
-        }
-    };
-    let repo = match db
-        .onboarded_targets()
+    let repo = match agent
+        .db
+        .repositories()
        .find_one(mongodb::bson::doc! { "_id": oid })
        .await
    {
        Ok(Some(repo)) => repo,
        _ => {
-            tracing::warn!("GitLab webhook: repo {repo_id} not found in tenant '{tenant_id}'");
+            tracing::warn!("GitLab webhook: repo {repo_id} not found");
            return StatusCode::NOT_FOUND;
        }
    };
@@ -65,21 +59,15 @@ pub async fn handle_gitlab_webhook(
        "push" => {
            let agent_clone = (*agent).clone();
            let repo_id = repo_id.clone();
-            let tenant_id = tenant_id.clone();
            tokio::spawn(async move {
-                tracing::info!(
-                    "GitLab push webhook: triggering scan for {repo_id} in tenant {tenant_id}"
-                );
-                if let Err(e) = agent_clone
-                    .run_scan(&tenant_id, &repo_id, ScanTrigger::Webhook)
-                    .await
-                {
+                tracing::info!("GitLab push webhook: triggering scan for {repo_id}");
+                if let Err(e) = agent_clone.run_scan(&repo_id, ScanTrigger::Webhook).await {
                    tracing::error!("Webhook-triggered scan failed: {e}");
                }
            });
            StatusCode::OK
        }
-        "merge_request" => handle_merge_request(agent, &tenant_id, &repo_id, &payload).await,
+        "merge_request" => handle_merge_request(agent, &repo_id, &payload).await,
        _ => {
            tracing::debug!("GitLab webhook: ignoring event '{event_type}'");
            StatusCode::OK
@@ -89,7 +77,6 @@ pub async fn handle_gitlab_webhook(

 async fn handle_merge_request(
    agent: Arc<ComplianceAgent>,
-    tenant_id: &str,
    repo_id: &str,
    payload: &serde_json::Value,
 ) -> StatusCode {
@@ -114,14 +101,13 @@ async fn handle_merge_request(
    }

    let repo_id = repo_id.to_string();
-    let tenant_id = tenant_id.to_string();
    let head_sha = head_sha.to_string();
    let base_sha = base_sha.to_string();
    let agent_clone = (*agent).clone();
    tokio::spawn(async move {
        tracing::info!("GitLab MR webhook: reviewing MR !{mr_iid} on {repo_id}");
        if let Err(e) = agent_clone
-            .run_pr_review(&tenant_id, &repo_id, mr_iid, &base_sha, &head_sha)
+            .run_pr_review(&repo_id, mr_iid, &base_sha, &head_sha)
            .await
        {
            tracing::error!("MR review failed for !{mr_iid}: {e}");
@@ -9,21 +9,17 @@ use crate::webhooks::{gitea, github, gitlab};

 pub async fn start_webhook_server(agent: &ComplianceAgent) -> Result<(), AgentError> {
    let app = Router::new()
-        // Per-tenant per-repo webhook URLs: /webhook/{tenant_id}/{platform}/{repo_id}
-        // The tenant_id is resolved from the URL path because webhooks
-        // arrive without a JWT — they're authenticated via per-repo HMAC,
-        // not via the tenant gate. The dashboard surfaces the full URL
-        // including the tenant_id when the repo is registered.
+        // Per-repo webhook URLs: /webhook/{platform}/{repo_id}
        .route(
-            "/webhook/{tenant_id}/github/{repo_id}",
+            "/webhook/github/{repo_id}",
            post(github::handle_github_webhook),
        )
        .route(
-            "/webhook/{tenant_id}/gitlab/{repo_id}",
+            "/webhook/gitlab/{repo_id}",
            post(gitlab::handle_gitlab_webhook),
        )
        .route(
-            "/webhook/{tenant_id}/gitea/{repo_id}",
+            "/webhook/gitea/{repo_id}",
            post(gitea::handle_gitea_webhook),
        )
        .layer(Extension(Arc::new(agent.clone())));
@@ -1,10 +0,0 @@
-//! Werkbank control-plane: the dynamic-execution job queue.
-//!
-//! The control plane enqueues declarative [`Job`](compliance_core::models::werkbank::Job)s
-//! and Werkbank runners lease, run, and complete them. [`queue::JobQueue`] is the
-//! Mongo-backed queue behind that flow (WB-02); the runner-facing HTTP transport
-//! and the runner itself land in later stories.
-
-pub mod queue;
-
-pub use queue::{JobQueue, SweepOutcome};
@@ -1,309 +0,0 @@
-//! The Mongo-backed Werkbank job queue (WB-02).
-//!
-//! A pull queue: the control plane [`enqueue`](JobQueue::enqueue)s jobs; a runner
-//! [`lease`](JobQueue::lease)s the oldest queued job it can run (matched by
-//! executor + labels), [`heartbeat`](JobQueue::heartbeat)s while it works, and
-//! [`complete`](JobQueue::complete)s it. Leases carry a visibility timeout: if a
-//! runner dies mid-job its heartbeats stop, the lease expires, and
-//! [`sweep_expired`](JobQueue::sweep_expired) returns the job to `queued` (or
-//! `expired` once it has been retried too many times).
-//!
-//! All state transitions are single atomic Mongo updates guarded by the lease
-//! token, so two runners can never both own a job. Every operation takes an
-//! explicit `now` so the queue's time-dependent behaviour is deterministically
-//! testable.
-
-use std::time::Duration;
-
-use chrono::{DateTime, Utc};
-use mongodb::bson::{doc, Bson, DateTime as BsonDateTime};
-use mongodb::error::{ErrorKind, WriteFailure};
-use mongodb::options::ReturnDocument;
-use mongodb::Collection;
-
-use compliance_core::models::werkbank::{
-    Executor, HeartbeatAck, Job, JobRecord, JobResult, JobStatus, LeasedJob,
-};
-
-use crate::database::Database;
-use crate::error::AgentError;
-
-/// The non-terminal states a job can be swept or cancelled from.
-const ACTIVE_STATES: [&str; 2] = ["leased", "running"];
-/// Every terminal state (no further transitions).
-const TERMINAL_STATES: [&str; 4] = ["succeeded", "failed", "expired", "cancelled"];
-
-/// What a visibility-timeout sweep did.
-#[derive(Debug, Clone, Copy, Default, PartialEq, Eq)]
-pub struct SweepOutcome {
-    /// Expired-lease jobs returned to `queued` for another runner.
-    pub requeued: u64,
-    /// Jobs that had exhausted their attempts and were marked `expired`.
-    pub expired: u64,
-}
-
-/// The Mongo-backed job queue.
-pub struct JobQueue {
-    coll: Collection<JobRecord>,
-}
-
-impl JobQueue {
-    /// Build a queue over a tenant database's `werkbank_jobs` collection.
-    pub fn new(db: &Database) -> Self {
-        Self {
-            coll: db.werkbank_jobs(),
-        }
-    }
-
-    /// Enqueue a job. Idempotent by job id: a job that is already present is a
-    /// no-op. Returns `true` if this call inserted it, `false` if it existed.
-    pub async fn enqueue(&self, job: Job, now: DateTime<Utc>) -> Result<bool, AgentError> {
-        let record = JobRecord::queued(job, now);
-        match self.coll.insert_one(&record).await {
-            Ok(_) => Ok(true),
-            Err(e) if is_duplicate_key(&e) => Ok(false),
-            Err(e) => Err(e.into()),
-        }
-    }
-
-    /// Atomically lease the oldest `queued` job this runner can run — matched by
-    /// executor and by labels (every label the job requires must be one the
-    /// runner advertises). Returns the job plus a lease token, or `None` if
-    /// nothing is runnable.
-    pub async fn lease(
-        &self,
-        runner_id: &str,
-        executor: Executor,
-        runner_labels: &[String],
-        lease_ttl: Duration,
-        now: DateTime<Utc>,
-    ) -> Result<Option<LeasedJob>, AgentError> {
-        let token = uuid::Uuid::new_v4().to_string();
-        let expires = bson_dt(now + ttl(lease_ttl));
-        let executor_bson = mongodb::bson::to_bson(&executor).unwrap_or(Bson::Null);
-
-        let filter = doc! {
-            "status": "queued",
-            "cancel_requested": { "$ne": true },
-            "job.executor": executor_bson,
-            // Every label the job requires must be in the runner's set — i.e. the
-            // job has no label that is not offered by the runner. Absent/empty
-            // job labels match any runner.
-            "job.labels": { "$not": { "$elemMatch": { "$nin": runner_labels.to_vec() } } },
-        };
-        let update = doc! {
-            "$set": {
-                "status": "leased",
-                "lease_token": &token,
-                "leased_by": runner_id,
-                "lease_expires_at": expires,
-                "heartbeat_at": bson_dt(now),
-                "updated_at": bson_dt(now),
-            },
-            "$inc": { "attempts": 1 },
-        };
-
-        let record = self
-            .coll
-            .find_one_and_update(filter, update)
-            .sort(doc! { "created_at": 1 }) // FIFO
-            .return_document(ReturnDocument::After)
-            .await?;
-        Ok(record.map(|r| LeasedJob {
-            job: r.job,
-            lease_token: token,
-        }))
-    }
-
-    /// Extend a lease and report whether the job has been asked to cancel.
-    /// Transitions the job to `running` on the first heartbeat. Returns `None`
-    /// when the lease is no longer valid (token mismatch, or the job is already
-    /// terminal) — the runner should then abandon the work.
-    pub async fn heartbeat(
-        &self,
-        job_id: &str,
-        lease_token: &str,
-        lease_ttl: Duration,
-        now: DateTime<Utc>,
-    ) -> Result<Option<HeartbeatAck>, AgentError> {
-        let filter = doc! {
-            "job.id": job_id,
-            "lease_token": lease_token,
-            "status": { "$in": ACTIVE_STATES.to_vec() },
-        };
-        let update = doc! {
-            "$set": {
-                "status": "running",
-                "lease_expires_at": bson_dt(now + ttl(lease_ttl)),
-                "heartbeat_at": bson_dt(now),
-                "updated_at": bson_dt(now),
-            },
-        };
-        let record = self
-            .coll
-            .find_one_and_update(filter, update)
-            .return_document(ReturnDocument::After)
-            .await?;
-        Ok(record.map(|r| HeartbeatAck {
-            cancelled: r.cancel_requested,
-        }))
-    }
-
-    /// Record a job's terminal result. Guarded by the lease token and only from
-    /// an active (`leased`/`running`) state, so it is idempotent — a duplicate or
-    /// late submission after the job already finished matches nothing. Returns
-    /// `true` if this call recorded the result.
-    pub async fn complete(
-        &self,
-        job_id: &str,
-        lease_token: &str,
-        result: &JobResult,
-        now: DateTime<Utc>,
-    ) -> Result<bool, AgentError> {
-        let status = result.status.unwrap_or(JobStatus::Failed);
-        let status_bson = mongodb::bson::to_bson(&status).unwrap_or(Bson::String("failed".into()));
-        let result_bson =
-            mongodb::bson::to_bson(result).map_err(|e| AgentError::Other(e.to_string()))?;
-
-        let filter = doc! {
-            "job.id": job_id,
-            "lease_token": lease_token,
-            "status": { "$in": ACTIVE_STATES.to_vec() },
-        };
-        let update = doc! {
-            "$set": {
-                "status": status_bson,
-                "result": result_bson,
-                "lease_token": Bson::Null,
-                "lease_expires_at": Bson::Null,
-                "updated_at": bson_dt(now),
-            },
-        };
-        let res = self.coll.update_one(filter, update).await?;
-        Ok(res.modified_count == 1)
-    }
-
-    /// Request cancellation of a job. A still-`queued` job is cancelled outright;
-    /// an in-flight one is flagged so the runner sees it on its next heartbeat and
-    /// tears down. Returns `true` if a non-terminal job matched.
-    pub async fn cancel(&self, job_id: &str, now: DateTime<Utc>) -> Result<bool, AgentError> {
-        let filter = doc! {
-            "job.id": job_id,
-            "status": { "$nin": TERMINAL_STATES.to_vec() },
-        };
-        // Pipeline update: flag cancellation, and if still queued flip straight to
-        // cancelled (nothing is running it).
-        let pipeline = vec![doc! {
-            "$set": {
-                "cancel_requested": true,
-                "status": {
-                    "$cond": [ { "$eq": ["$status", "queued"] }, "cancelled", "$status" ]
-                },
-                "updated_at": bson_dt(now),
-            }
-        }];
-        let res = self.coll.update_one(filter, pipeline).await?;
-        Ok(res.matched_count == 1)
-    }
-
-    /// Sweep leases whose visibility timeout has elapsed: return them to `queued`
-    /// for another runner, or mark them `expired` once they have been leased
-    /// `max_attempts` times. This is what makes a crashed runner's job recover.
-    pub async fn sweep_expired(
-        &self,
-        now: DateTime<Utc>,
-        max_attempts: u32,
-        // (kept explicit rather than a const so callers can tune retry policy)
-    ) -> Result<SweepOutcome, AgentError> {
-        let now_bson = bson_dt(now);
-        let max = i64::from(max_attempts);
-
-        let requeue = self
-            .coll
-            .update_many(
-                doc! {
-                    "status": { "$in": ACTIVE_STATES.to_vec() },
-                    "lease_expires_at": { "$lt": &now_bson },
-                    "attempts": { "$lt": max },
-                },
-                doc! { "$set": {
-                    "status": "queued",
-                    "lease_token": Bson::Null,
-                    "leased_by": Bson::Null,
-                    "lease_expires_at": Bson::Null,
-                    "updated_at": &now_bson,
-                } },
-            )
-            .await?;
-
-        let expire = self
-            .coll
-            .update_many(
-                doc! {
-                    "status": { "$in": ACTIVE_STATES.to_vec() },
-                    "lease_expires_at": { "$lt": &now_bson },
-                    "attempts": { "$gte": max },
-                },
-                doc! { "$set": {
-                    "status": "expired",
-                    "lease_token": Bson::Null,
-                    "lease_expires_at": Bson::Null,
-                    "updated_at": &now_bson,
-                } },
-            )
-            .await?;
-
-        Ok(SweepOutcome {
-            requeued: requeue.modified_count,
-            expired: expire.modified_count,
-        })
-    }
-
-    /// Fetch a job record by job id (inspection / control-plane reads).
-    pub async fn get(&self, job_id: &str) -> Result<Option<JobRecord>, AgentError> {
-        Ok(self.coll.find_one(doc! { "job.id": job_id }).await?)
-    }
-}
-
-/// A `chrono::Duration` for a lease TTL, saturating rather than panicking on an
-/// absurd input (`chrono::Duration::seconds` panics past its internal bound).
-fn ttl(d: Duration) -> chrono::Duration {
-    let secs = i64::try_from(d.as_secs()).unwrap_or(i64::MAX);
-    chrono::Duration::try_seconds(secs).unwrap_or(chrono::Duration::MAX)
-}
-
-/// A chrono instant as a BSON date (so Mongo stores/compares it as a real date).
-fn bson_dt(dt: DateTime<Utc>) -> BsonDateTime {
-    BsonDateTime::from_chrono(dt)
-}
-
-/// Whether a Mongo error is a duplicate-key (E11000) violation — a job with this
-/// id is already enqueued.
-fn is_duplicate_key(e: &mongodb::error::Error) -> bool {
-    match &*e.kind {
-        ErrorKind::Write(WriteFailure::WriteError(we)) => we.code == 11000,
-        _ => false,
-    }
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    #[test]
-    fn ttl_saturates_and_converts() {
-        assert_eq!(ttl(Duration::from_secs(30)), chrono::Duration::seconds(30));
-        // An absurd TTL saturates instead of panicking.
-        assert_eq!(ttl(Duration::from_secs(u64::MAX)), chrono::Duration::MAX);
-    }
-
-    #[test]
-    fn state_constants_are_disjoint() {
-        for s in ACTIVE_STATES {
-            assert!(
-                !TERMINAL_STATES.contains(&s),
-                "{s} cannot be both active and terminal"
-            );
-        }
-    }
-}
@@ -1,125 +0,0 @@
-//! C5 example 2 — exploratory (not a committed regression test). Four topically
-//! distinct findings, to see whether tuned semantic retrieval maps each to the
-//! right master-control family. Run:
-//!   export ... (LITELLM_* + BREAKPILOT_BASE_URL)
-//!   cargo test -p compliance-agent --test c5_example2 -- --ignored --nocapture
-
-mod common;
-
-use std::sync::Arc;
-
-use compliance_agent::llm::LlmClient;
-use compliance_core::config::BreakpilotConfig;
-use compliance_core::models::finding::{Finding, Severity};
-use compliance_core::models::scan::ScanType;
-use secrecy::SecretString;
-
-fn env(k: &str) -> String {
-    std::env::var(k).unwrap_or_else(|_| panic!("env {k} must be set"))
-}
-
-fn mk(file: &str, line: u32, title: &str, desc: &str) -> Finding {
-    let mut f = Finding::new(
-        "repo-c5b".into(),
-        format!("{file}:{line}"),
-        "semgrep".into(),
-        ScanType::Sast,
-        title.into(),
-        desc.into(),
-        Severity::High,
-    );
-    f.file_path = Some(file.into());
-    f.line_number = Some(line);
-    f
-}
-
-fn write(repo: &std::path::Path, rel: &str, body: &str) {
-    let p = repo.join(rel);
-    if let Some(parent) = p.parent() {
-        std::fs::create_dir_all(parent).unwrap();
-    }
-    std::fs::write(p, body).unwrap();
-}
-
-#[tokio::test]
-#[ignore = "live: api-dev + LiteLLM"]
-async fn c5b_varied_findings() {
-    let llm = Arc::new(LlmClient::new(
-        env("LITELLM_URL"),
-        SecretString::from(env("LITELLM_API_KEY")),
-        env("LITELLM_MODEL"),
-        env("LITELLM_EMBED_MODEL"),
-    ));
-    let mut config = common::dev_config("mongodb://unused".into(), "c5b".into());
-    config.breakpilot = BreakpilotConfig {
-        base_url: Some(env("BREAKPILOT_BASE_URL")),
-        token: None,
-        snapshot_dir: std::env::temp_dir()
-            .join("c5-oscal-snap")
-            .to_string_lossy()
-            .into_owned(),
-        semantic_mapping: true,
-        grounded_control_checks: false,
-    };
-
-    let repo = std::env::temp_dir().join("c5b-fixture-repo");
-    let _ = std::fs::remove_dir_all(&repo);
-    write(
-        &repo,
-        "app/db.py",
-        "import sqlite3\n\ndef get_user(username):\n    q = \"SELECT * FROM users WHERE name = '\" + username + \"'\"\n    return conn.execute(q)\n",
-    );
-    write(
-        &repo,
-        "app/config.py",
-        "# service config\nAPI_KEY = \"sk_live_51H8xYz3kQ9v2bNmR7wT4uSpQ\"\nDB_HOST = \"db.internal\"\n",
-    );
-    write(
-        &repo,
-        "app/net.py",
-        "import requests\n\ndef fetch(url):\n    return requests.get(url, verify=False, timeout=5)\n",
-    );
-    write(
-        &repo,
-        "app/ser.py",
-        "import pickle\n\ndef load_state(blob):\n    return pickle.loads(blob)\n",
-    );
-
-    let mut findings = vec![
-        mk(
-            "app/db.py",
-            4,
-            "SQL injection via string-concatenated query",
-            "User input is concatenated directly into a SQL statement, allowing SQL injection.",
-        ),
-        mk(
-            "app/config.py",
-            2,
-            "Hardcoded API credential in source",
-            "A live API key is hardcoded in source code instead of a secret store.",
-        ),
-        mk(
-            "app/net.py",
-            4,
-            "TLS certificate verification disabled",
-            "requests is called with verify=False, disabling TLS certificate validation.",
-        ),
-        mk(
-            "app/ser.py",
-            3,
-            "Insecure deserialization with pickle.loads",
-            "Untrusted data is deserialized with pickle.loads, allowing remote code execution.",
-        ),
-    ];
-
-    let tagged =
-        compliance_agent::controls::semantic_stamp_findings(&config, llm, &repo, &mut findings)
-            .await;
-    println!("\n=== C5 example 2: varied findings ===");
-    for f in &findings {
-        println!("  {:52} -> {:?}", f.title, f.control_refs);
-    }
-    println!("tagged: {tagged}/4");
-    let _ = std::fs::remove_dir_all(&repo);
-    assert!(tagged >= 1);
-}
@@ -1,145 +0,0 @@
-//! C5 live verification — the semantic master-controls path end to end against the
-//! deployed api-dev catalog. Ignored (hits api-dev + LiteLLM). Run explicitly:
-//!
-//!   set -a; . ./.env; set +a
-//!   BREAKPILOT_BASE_URL=https://api-dev.breakpilot.ai \
-//!     cargo test -p compliance-agent --test c5_semantic_live -- --ignored --nocapture
-//!
-//! Pulls the live master-controls catalog, embeds the corpus (chunked), then for a
-//! couple of real vulnerable findings retrieves the nearest master controls and
-//! grounded-judges them, stamping master-control refs.
-
-mod common;
-
-use std::sync::Arc;
-
-use compliance_agent::llm::LlmClient;
-use compliance_core::config::BreakpilotConfig;
-use compliance_core::models::finding::{Finding, Severity};
-use compliance_core::models::scan::ScanType;
-use secrecy::SecretString;
-
-fn env(k: &str) -> String {
-    std::env::var(k).unwrap_or_else(|_| panic!("env {k} must be set for the live C5 test"))
-}
-
-fn mk_finding(file: &str, line: u32, title: &str) -> Finding {
-    let mut f = Finding::new(
-        "repo-c5".into(),
-        format!("{file}:{line}"),
-        "semgrep".into(),
-        ScanType::Sast,
-        title.into(),
-        title.into(),
-        Severity::High,
-    );
-    f.file_path = Some(file.into());
-    f.line_number = Some(line);
-    f
-}
-
-#[tokio::test]
-#[ignore = "live: requires deployed api-dev master-controls (fetch+parse only, no LLM)"]
-async fn c5_ingest_master_controls_catalog() {
-    use compliance_agent::controls::OscalControlsProvider;
-
-    let provider = OscalControlsProvider::new(
-        reqwest::Client::new(),
-        env("BREAKPILOT_BASE_URL"),
-        None,
-        std::env::temp_dir().join("c5-ingest-snap"),
-    );
-    let doc = provider
-        .load_master_controls()
-        .await
-        .expect("pull + parse master-controls catalog");
-    let controls = doc.to_controls();
-    println!(
-        "\n=== C5 ingest: {} master controls parsed ===",
-        controls.len()
-    );
-    for c in controls.iter().take(4) {
-        let text: String = c.text.chars().take(90).collect();
-        println!("  {} | {} | {}", c.id, c.title, text);
-    }
-    assert!(
-        !controls.is_empty(),
-        "expected a non-empty master-control corpus"
-    );
-}
-
-#[tokio::test]
-#[ignore = "live: requires deployed api-dev master-controls + LiteLLM"]
-async fn c5_semantic_stamps_master_control_refs() {
-    let llm = Arc::new(LlmClient::new(
-        env("LITELLM_URL"),
-        SecretString::from(env("LITELLM_API_KEY")),
-        env("LITELLM_MODEL"),
-        env("LITELLM_EMBED_MODEL"),
-    ));
-
-    let mut config = common::dev_config("mongodb://unused".into(), "c5".into());
-    let snapshot = std::env::temp_dir().join("c5-oscal-snap");
-    config.breakpilot = BreakpilotConfig {
-        base_url: Some(env("BREAKPILOT_BASE_URL")),
-        token: None,
-        snapshot_dir: snapshot.to_string_lossy().into_owned(),
-        semantic_mapping: true,
-        grounded_control_checks: false,
-    };
-
-    // Fixture repo with recognizable code-checkable surfaces.
-    let repo = std::env::temp_dir().join("c5-fixture-repo");
-    let _ = std::fs::remove_dir_all(&repo);
-    std::fs::create_dir_all(repo.join("app")).expect("mkdir");
-    std::fs::write(
-        repo.join("app/auth.py"),
-        concat!(
-            "import hashlib\n",
-            "\n",
-            "def store_password(user, password):\n",
-            "    # weak, unsalted password hashing\n",
-            "    digest = hashlib.md5(password.encode()).hexdigest()\n",
-            "    db.save(user, digest)\n",
-            "\n",
-            "@app.route('/login', methods=['POST'])\n",
-            "def login():\n",
-            "    u = request.form['username']\n",
-            "    p = request.form['password']\n",
-            "    return 'ok' if check(u, p) else ('bad', 401)\n",
-        ),
-    )
-    .expect("write fixture");
-
-    let mut findings = vec![
-        mk_finding("app/auth.py", 5, "Weak password hash (md5, unsalted)"),
-        mk_finding(
-            "app/auth.py",
-            9,
-            "Login endpoint without brute-force protection",
-        ),
-    ];
-
-    let tagged =
-        compliance_agent::controls::semantic_stamp_findings(&config, llm, &repo, &mut findings)
-            .await;
-
-    println!("\n=== C5 semantic master-controls stamping ===");
-    for f in &findings {
-        println!(
-            "  {:50} {}:{:?} -> {:?}",
-            f.title,
-            f.file_path.as_deref().unwrap_or(""),
-            f.line_number,
-            f.control_refs
-        );
-    }
-    println!("findings that gained >=1 master-control ref: {tagged}");
-    let _ = std::fs::remove_dir_all(&repo);
-
-    // Live corpus — assert only that the path runs and stamps at least one ref.
-    assert!(
-        tagged >= 1,
-        "expected at least one finding to gain a master-control ref"
-    );
-}
@@ -1,197 +1,3 @@
-// Shared test harness for E2E / integration tests.
+// Shared test helpers for compliance-agent integration tests.
 //
-// Spins up the agent API server on a random port with an isolated test
-// database. Each test gets a fresh database that is dropped on cleanup.
-//
-// Included via `mod common;` in several test binaries; not every binary uses
-// every helper, so allow dead code here.
-#![allow(dead_code)]
-
-use std::sync::Arc;
-
-use compliance_agent::agent::ComplianceAgent;
-use compliance_agent::api;
-use compliance_agent::database::DatabasePool;
-use compliance_core::AgentConfig;
-use secrecy::SecretString;
-
-/// The runner bearer token wired into the test config.
-pub const TEST_RUNNER_TOKEN: &str = "test-runner-token";
-
-/// A minimal dev [`AgentConfig`] for tests: unauthenticated (no Keycloak), the
-/// Werkbank runner API enabled with [`TEST_RUNNER_TOKEN`].
-pub fn dev_config(mongodb_uri: String, db_name: String) -> AgentConfig {
-    AgentConfig {
-        mongodb_uri,
-        mongodb_database: db_name,
-        litellm_url: std::env::var("TEST_LITELLM_URL")
-            .unwrap_or_else(|_| "http://localhost:4000".into()),
-        litellm_api_key: SecretString::from(String::new()),
-        litellm_model: "gpt-4o".into(),
-        litellm_embed_model: "text-embedding-3-small".into(),
-        agent_port: 0, // not used — we bind ourselves
-        scan_schedule: String::new(),
-        cve_monitor_schedule: String::new(),
-        git_clone_base_path: "/tmp/compliance-scanner-tests/repos".into(),
-        artifact_store_base_path: "/tmp/compliance-scanner-tests/artifacts".into(),
-        ssh_key_path: "/tmp/compliance-scanner-tests/ssh/id_ed25519".into(),
-        github_token: None,
-        github_webhook_secret: None,
-        gitlab_url: None,
-        gitlab_token: None,
-        gitlab_webhook_secret: None,
-        jira_url: None,
-        jira_email: None,
-        jira_api_token: None,
-        jira_project_key: None,
-        searxng_url: None,
-        nvd_api_key: None,
-        keycloak_url: None,
-        keycloak_realm: None,
-        keycloak_admin_username: None,
-        keycloak_admin_password: None,
-        pentest_verification_email: None,
-        pentest_imap_host: None,
-        pentest_imap_port: None,
-        pentest_imap_tls: false,
-        pentest_imap_username: None,
-        pentest_imap_password: None,
-        admin_api_token: None,
-        tenant_registry_url: None,
-        plc_runtime: compliance_core::PlcRuntimeConfig::default(),
-        werkbank_runner_token: Some(SecretString::from(TEST_RUNNER_TOKEN.to_string())),
-        breakpilot: compliance_core::config::BreakpilotConfig::default(),
-    }
-}
-
-/// A running test server with a unique database.
-pub struct TestServer {
-    pub base_url: String,
-    pub client: reqwest::Client,
-    db_name: String,
-    mongodb_uri: String,
-}
-
-impl TestServer {
-    /// Start an agent API server on a random port with an isolated database.
-    pub async fn start() -> Self {
-        let mongodb_uri = std::env::var("TEST_MONGODB_URI")
-            .unwrap_or_else(|_| "mongodb://root:example@localhost:27017/?authSource=admin".into());
-
-        // Unique db-name prefix per run. Must fit the pool's 30-char cap
-        // (`<prefix>_<32 hex>` <= 63), so use a 16-hex-char suffix.
-        let db_name = format!("t_{}", &uuid::Uuid::new_v4().simple().to_string()[..16]);
-
-        let db_pool = DatabasePool::connect(&mongodb_uri, &db_name)
-            .await
-            .expect("Failed to build DatabasePool");
-
-        let config = dev_config(mongodb_uri.clone(), db_name.clone());
-
-        let agent = ComplianceAgent::new(config, db_pool);
-
-        // Build the router with the agent extension. After M7.2-B every
-        // handler takes a TenantCtx extractor; without KC in the test
-        // harness, the dev-tenant injector mounts a synthetic context so
-        // tests run end-to-end against `<db_name>_dev`.
-        let app = api::routes::build_router()
-            .layer(axum::extract::Extension(Arc::new(agent)))
-            .layer(axum::middleware::from_fn(api::server::inject_dev_tenant))
-            .layer(tower_http::cors::CorsLayer::permissive());
-
-        // Bind to port 0 to get a random available port
-        let listener = tokio::net::TcpListener::bind("127.0.0.1:0")
-            .await
-            .expect("Failed to bind test server");
-        let port = listener.local_addr().expect("no local addr").port();
-
-        tokio::spawn(async move {
-            axum::serve(listener, app).await.ok();
-        });
-
-        let base_url = format!("http://127.0.0.1:{port}");
-        let client = reqwest::Client::builder()
-            .timeout(std::time::Duration::from_secs(30))
-            .build()
-            .expect("Failed to build HTTP client");
-
-        // Wait for server to be ready
-        for _ in 0..50 {
-            if client
-                .get(format!("{base_url}/api/v1/health"))
-                .send()
-                .await
-                .is_ok()
-            {
-                break;
-            }
-            tokio::time::sleep(std::time::Duration::from_millis(50)).await;
-        }
-
-        Self {
-            base_url,
-            client,
-            db_name,
-            mongodb_uri,
-        }
-    }
-
-    /// GET helper
-    pub async fn get(&self, path: &str) -> reqwest::Response {
-        self.client
-            .get(format!("{}{path}", self.base_url))
-            .send()
-            .await
-            .expect("GET request failed")
-    }
-
-    /// POST helper with JSON body
-    pub async fn post(&self, path: &str, body: &serde_json::Value) -> reqwest::Response {
-        self.client
-            .post(format!("{}{path}", self.base_url))
-            .json(body)
-            .send()
-            .await
-            .expect("POST request failed")
-    }
-
-    /// PATCH helper with JSON body
-    pub async fn patch(&self, path: &str, body: &serde_json::Value) -> reqwest::Response {
-        self.client
-            .patch(format!("{}{path}", self.base_url))
-            .json(body)
-            .send()
-            .await
-            .expect("PATCH request failed")
-    }
-
-    /// DELETE helper
-    pub async fn delete(&self, path: &str) -> reqwest::Response {
-        self.client
-            .delete(format!("{}{path}", self.base_url))
-            .send()
-            .await
-            .expect("DELETE request failed")
-    }
-
-    /// Get the unique database name for direct MongoDB access in tests.
-    pub fn db_name(&self) -> &str {
-        &self.db_name
-    }
-
-    /// Drop every per-tenant database belonging to this test run.
-    /// Post-M7.2-D the agent never opens a `db_name` directly —
-    /// data lives only in `<db_name>_<tenant>` per-tenant databases.
-    pub async fn cleanup(&self) {
-        if let Ok(client) = mongodb::Client::with_uri_str(&self.mongodb_uri).await {
-            if let Ok(names) = client.list_database_names().await {
-                let prefix = format!("{}_", self.db_name);
-                for name in names {
-                    if name.starts_with(&prefix) {
-                        client.database(&name).drop().await.ok();
-                    }
-                }
-            }
-        }
-    }
-}
+// Add database mocks, fixtures, and test utilities here.
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
Sharang ParnerkarandClaude Opus 4.6	f1bb053882	feat: deduplicate code review findings across LLM passes Group findings by file, line proximity, and normalized title keywords, keeping the highest-severity finding from each group and merging CWE info. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-25 21:55:52 +01:00
Sharang ParnerkarandClaude Opus 4.6	af4760baf5	fix: update rustls-webpki to 0.103.10 (RUSTSEC-2026-0049) CI / Deploy Dashboard (pull_request) Has been skipped Details CI / Deploy Docs (pull_request) Has been skipped Details CI / Deploy MCP (pull_request) Has been skipped Details CI / Check (pull_request) Successful in 13m17s Details CI / Detect Changes (pull_request) Has been skipped Details CI / Deploy Agent (pull_request) Has been skipped Details Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-25 17:08:18 +01:00
Sharang ParnerkarandClaude Opus 4.6	f310a3e0a2	style: apply cargo fmt formatting CI / Detect Changes (pull_request) Has been skipped Details CI / Check (pull_request) Failing after 6m17s Details CI / Deploy Docs (pull_request) Has been skipped Details CI / Deploy MCP (pull_request) Has been skipped Details CI / Deploy Agent (pull_request) Has been skipped Details CI / Deploy Dashboard (pull_request) Has been skipped Details Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-25 16:48:08 +01:00
Sharang ParnerkarandClaude Opus 4.6	e5c14636a7	fix: avoid unwrap in MCP health check to pass clippy CI / Check (pull_request) Failing after 2m52s Details CI / Detect Changes (pull_request) Has been skipped Details CI / Deploy Agent (pull_request) Has been skipped Details CI / Deploy Dashboard (pull_request) Has been skipped Details CI / Deploy Docs (pull_request) Has been skipped Details CI / Deploy MCP (pull_request) Has been skipped Details Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-25 16:32:34 +01:00
Sharang ParnerkarandClaude Opus 4.6	fad8bbbd65	feat: add health check for MCP servers on dashboard CI / Check (pull_request) Failing after 1m44s Details CI / Detect Changes (pull_request) Has been skipped Details CI / Deploy Agent (pull_request) Has been skipped Details CI / Deploy Dashboard (pull_request) Has been skipped Details CI / Deploy Docs (pull_request) Has been skipped Details CI / Deploy MCP (pull_request) Has been skipped Details Add /health endpoint to the MCP server and probe it from the dashboard on page load to update server status (running/stopped) in real time. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-20 22:51:39 +01:00
Sharang ParnerkarandClaude Opus 4.6	c62e9fdcd4	fix: check Gitea API response status and fallback for PR reviews CI / Check (pull_request) Failing after 1m28s Details CI / Detect Changes (pull_request) Has been skipped Details CI / Deploy Agent (pull_request) Has been skipped Details CI / Deploy Dashboard (pull_request) Has been skipped Details CI / Deploy Docs (pull_request) Has been skipped Details CI / Deploy MCP (pull_request) Has been skipped Details The Gitea tracker was silently swallowing HTTP error responses — only network failures were caught, so 4xx/5xx from Gitea went unnoticed and the agent logged success even when reviews weren't posted. - Add response status checking to create_pr_review, update_issue_status, and add_comment - Add fallback in create_pr_review: if inline comments fail, retry as a plain PR comment so the summary still gets posted Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>	2026-03-18 23:41:03 +01:00