(1) B22 Cross-Domain (fix #59):
Elli-Test fand AGB auf logpay.de NICHT obwohl URL in doc_entries
korrekt. Vermutete Ursache: Discovery-Phase A drops/überschreibt
Original-URL bei PDF-Fetch-Fail (word_count=0).
Fix: _collect_audit_urls() iteriert über state.doc_entries +
rejected_url + req.documents — Cross-Domain-Hosting ist
unabhängig vom Text-Inhalt. Plus Trace-Logging für künftige
Diagnose. Dedup per (doc_type, host_sld).
(2) B17 Audit-Walk-Fail-Fallback (fix #60):
BMW v5 hatte audit_walk=None ohne Mail-Hinweis. Vermutlich
180s-Timeout bei OneTrust-CMP-Banner-Tour.
Fix: Timeout 180s → 300s. Plus: Bei Fail wird ein Hinweis-
Stub mit error-Grund in state["audit_walk"] + HTML-Block
geschrieben — Reviewer sieht den Fail statt silent-skip.
(3) company_name + origin_domain im Backend (fix #61):
Frontend sendet seit ec03317 die zwei Felder — Backend ignorierte
sie.
Fix: ComplianceCheckRequest-Schema um company_name +
origin_domain erweitert. phase_e_email priorisiert User-Input
vor URL-Heuristik für site_name. Bei origin_domain ohne
ableitbare doc_entries-domain wird der User-Input als domain
übernommen.
(4) Plausibility-LLM Fallback-Modell (fix #62):
qwen3:30b-a3b liefert auf großen DSEs (BMW 122 FAIL) gehäuft
leere format='json'-Responses — Circuit-Breaker griff aber
Phase blieb nutzlos.
Fix: Default-Modell auf qwen2.5:7b umgestellt (4× kleiner,
zuverlässiger bei format=json, ausreichendes Reasoning für
PASS/MODIFY/DROP-Klassifikation). Plus Strategy-C eingeführt
— Fallback-Modell (llama3.2:3b) wenn primary leer bleibt.
BATCH_SIZE 4 → 3. ENV-Switches PLAUSIBILITY_LLM_MODEL +
PLAUSIBILITY_FALLBACK_MODEL für Tuning.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
breakpilot-compliance
DSGVO/AI-Act compliance platform — 10 services, Go · Python · TypeScript
Overview
breakpilot-compliance is a multi-tenant DSGVO/EU AI Act compliance platform that provides an SDK for consent management, data subject requests (DSR), audit logging, iACE impact assessments, and document archival. It ships as 10 containerised services covering an admin dashboard, a developer portal, a Python/FastAPI backend, a Go AI compliance engine, TTS, and a decentralised document store on IPFS. Every service is deployed automatically via Gitea Actions → Orca on every push to main.
Architecture
| Service | Tech | Port | Container |
|---|---|---|---|
| admin-compliance | Next.js 15 | 3007 | bp-compliance-admin |
| backend-compliance | Python / FastAPI 0.123 | 8002 | bp-compliance-backend |
| ai-compliance-sdk | Go 1.24 / Gin | 8093 | bp-compliance-ai-sdk |
| developer-portal | Next.js 15 | 3006 | bp-compliance-developer-portal |
| breakpilot-compliance-sdk | TypeScript SDK (React/Vue/Angular/vanilla) | — | — |
| consent-sdk | JS/TS Consent SDK | — | — |
| compliance-tts-service | Python / Piper TTS | 8095 | bp-compliance-tts |
| document-crawler | Python / FastAPI | 8098 | bp-compliance-document-crawler |
| dsms-gateway | Python / FastAPI / IPFS | 8082 | bp-compliance-dsms-gateway |
| dsms-node | IPFS Kubo v0.24.0 | — | bp-compliance-dsms-node |
All containers share the external breakpilot-network Docker network and depend on breakpilot-core (Valkey, Vault, RAG service, Nginx reverse proxy).
Quick Start
Prerequisites: Docker, Go 1.24+, Python 3.12+, Node.js 20+
git clone ssh://git@gitea.meghsakha.com:22222/Benjamin_Boenisch/breakpilot-compliance.git
cd breakpilot-compliance
# Copy and populate secrets (never commit .env)
cp .env.example .env
# Start all services
docker compose up -d
For the Orca/Hetzner production target (x86_64), use the override:
docker compose -f docker-compose.yml -f docker-compose.hetzner.yml up -d
Development Workflow
Use feature branches off main. Supported prefixes: feat/, feature/, hotfix/.
git checkout main && git pull origin main
git checkout -b feat/my-change
# ... make changes ...
git push origin feat/my-change
# Open a PR → squash merge to main
Push to main triggers:
- Gitea Actions — lint → test → validate (see CI Pipeline below)
- Orca — automatic build + deploy (~3 min total)
Monitor status: https://gitea.meghsakha.com/Benjamin_Boenisch/breakpilot-compliance/actions
CI Pipeline
Defined in .gitea/workflows/ci.yaml.
| Job | What it checks |
|---|---|
loc-budget |
All source files ≤ 500 LOC; soft target 300 |
guardrail-integrity |
Commits touching guardrail files carry [guardrail-change] |
go-lint |
golangci-lint on ai-compliance-sdk/ |
python-lint |
ruff + mypy on Python services |
nodejs-lint |
tsc --noEmit + ESLint on Next.js services |
test-go-ai-compliance |
go test ./... in ai-compliance-sdk/ |
test-python-backend-compliance |
pytest in backend-compliance/ |
test-python-document-crawler |
pytest in document-crawler/ |
test-python-dsms-gateway |
pytest test_main.py in dsms-gateway/ |
sbom-scan |
License + vulnerability scan via syft + grype |
validate-canonical-controls |
OpenAPI contract baseline diff |
File Budget
| Limit | Value | How to check |
|---|---|---|
| Soft target | 300 LOC | bash scripts/check-loc.sh |
| Hard cap | 500 LOC | Same; also enforced by PreToolUse hook + git pre-commit + CI |
| Exceptions | .claude/rules/loc-exceptions.txt |
Require written rationale + [guardrail-change] commit marker |
The .claude/settings.json PreToolUse hook blocks Claude Code from writing or editing files that would exceed the hard cap. The git pre-commit hook re-checks. CI is the final gate.
Links
| URL | |
|---|---|
| Admin dashboard | https://admin-dev.breakpilot.ai |
| Developer portal | https://developers-dev.breakpilot.ai |
| Backend API | https://api-dev.breakpilot.ai |
| AI SDK API | https://sdk-dev.breakpilot.ai |
| Gitea repo | https://gitea.meghsakha.com/Benjamin_Boenisch/breakpilot-compliance |
| Gitea Actions | https://gitea.meghsakha.com/Benjamin_Boenisch/breakpilot-compliance/actions |