feat: enhance tracing with field attributes and warn logging across all handlers
All checks were successful
CI / Tests (push) Successful in 5m17s
CI / Detect Changes (push) Successful in 3s
CI / Deploy Agent (push) Successful in 3s
CI / Deploy Dashboard (push) Has been skipped
CI / Deploy Docs (push) Has been skipped
CI / Deploy MCP (push) Has been skipped
CI / Format (push) Successful in 4s
CI / Clippy (push) Successful in 4m38s
CI / Security Audit (push) Successful in 1m50s

Add repo_id, finding_id, and filter fields to tracing::instrument attributes
for better trace correlation in SigNoz. Replace all silently swallowed errors
(Err(_) => Vec::new()) with tracing::warn! logging across mod.rs, dast.rs,
graph.rs handlers. Add stage-level spans with .instrument() to pipeline
orchestrator for visibility into scan phases.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Sharang Parnerkar
2026-03-10 21:56:16 +01:00
parent 67d6a937ae
commit 99983c51e3
8 changed files with 178 additions and 70 deletions

View File

@@ -17,7 +17,7 @@ use super::ApiResponse;
type AgentExt = Extension<Arc<ComplianceAgent>>; type AgentExt = Extension<Arc<ComplianceAgent>>;
/// POST /api/v1/chat/:repo_id — Send a chat message with RAG context /// POST /api/v1/chat/:repo_id — Send a chat message with RAG context
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
pub async fn chat( pub async fn chat(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(repo_id): Path<String>, Path(repo_id): Path<String>,
@@ -127,7 +127,7 @@ pub async fn chat(
} }
/// POST /api/v1/chat/:repo_id/build-embeddings — Trigger embedding build /// POST /api/v1/chat/:repo_id/build-embeddings — Trigger embedding build
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
pub async fn build_embeddings( pub async fn build_embeddings(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(repo_id): Path<String>, Path(repo_id): Path<String>,
@@ -228,7 +228,7 @@ pub async fn build_embeddings(
} }
/// GET /api/v1/chat/:repo_id/status — Get latest embedding build status /// GET /api/v1/chat/:repo_id/status — Get latest embedding build status
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
pub async fn embedding_status( pub async fn embedding_status(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(repo_id): Path<String>, Path(repo_id): Path<String>,

View File

@@ -63,7 +63,10 @@ pub async fn list_targets(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch DAST targets: {e}");
Vec::new()
}
}; };
Ok(Json(ApiResponse { Ok(Json(ApiResponse {
@@ -101,7 +104,7 @@ pub async fn add_target(
} }
/// POST /api/v1/dast/targets/:id/scan — Trigger DAST scan /// POST /api/v1/dast/targets/:id/scan — Trigger DAST scan
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(target_id = %id))]
pub async fn trigger_scan( pub async fn trigger_scan(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(id): Path<String>, Path(id): Path<String>,
@@ -163,7 +166,10 @@ pub async fn list_scan_runs(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch DAST scan runs: {e}");
Vec::new()
}
}; };
Ok(Json(ApiResponse { Ok(Json(ApiResponse {
@@ -196,7 +202,10 @@ pub async fn list_findings(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch DAST findings: {e}");
Vec::new()
}
}; };
Ok(Json(ApiResponse { Ok(Json(ApiResponse {
@@ -207,7 +216,7 @@ pub async fn list_findings(
} }
/// GET /api/v1/dast/findings/:id — Finding detail with evidence /// GET /api/v1/dast/findings/:id — Finding detail with evidence
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(finding_id = %id))]
pub async fn get_finding( pub async fn get_finding(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(id): Path<String>, Path(id): Path<String>,

View File

@@ -33,7 +33,7 @@ fn default_search_limit() -> usize {
} }
/// GET /api/v1/graph/:repo_id — Full graph data /// GET /api/v1/graph/:repo_id — Full graph data
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
pub async fn get_graph( pub async fn get_graph(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(repo_id): Path<String>, Path(repo_id): Path<String>,
@@ -55,11 +55,17 @@ pub async fn get_graph(
let all_nodes: Vec<CodeNode> = match db.graph_nodes().find(filter.clone()).await { let all_nodes: Vec<CodeNode> = match db.graph_nodes().find(filter.clone()).await {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch graph nodes: {e}");
Vec::new()
}
}; };
let edges: Vec<CodeEdge> = match db.graph_edges().find(filter).await { let edges: Vec<CodeEdge> = match db.graph_edges().find(filter).await {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch graph edges: {e}");
Vec::new()
}
}; };
// Remove disconnected nodes (no edges) to keep the graph clean // Remove disconnected nodes (no edges) to keep the graph clean
@@ -89,7 +95,7 @@ pub async fn get_graph(
} }
/// GET /api/v1/graph/:repo_id/nodes — List nodes (paginated) /// GET /api/v1/graph/:repo_id/nodes — List nodes (paginated)
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
pub async fn get_nodes( pub async fn get_nodes(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(repo_id): Path<String>, Path(repo_id): Path<String>,
@@ -99,7 +105,10 @@ pub async fn get_nodes(
let nodes: Vec<CodeNode> = match db.graph_nodes().find(filter).await { let nodes: Vec<CodeNode> = match db.graph_nodes().find(filter).await {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch graph nodes: {e}");
Vec::new()
}
}; };
let total = nodes.len() as u64; let total = nodes.len() as u64;
@@ -111,7 +120,7 @@ pub async fn get_nodes(
} }
/// GET /api/v1/graph/:repo_id/communities — List detected communities /// GET /api/v1/graph/:repo_id/communities — List detected communities
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
pub async fn get_communities( pub async fn get_communities(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(repo_id): Path<String>, Path(repo_id): Path<String>,
@@ -121,7 +130,10 @@ pub async fn get_communities(
let nodes: Vec<CodeNode> = match db.graph_nodes().find(filter).await { let nodes: Vec<CodeNode> = match db.graph_nodes().find(filter).await {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch graph nodes for communities: {e}");
Vec::new()
}
}; };
let mut communities: std::collections::HashMap<u32, Vec<String>> = let mut communities: std::collections::HashMap<u32, Vec<String>> =
@@ -161,7 +173,7 @@ pub struct CommunityInfo {
} }
/// GET /api/v1/graph/:repo_id/impact/:finding_id — Impact analysis /// GET /api/v1/graph/:repo_id/impact/:finding_id — Impact analysis
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %repo_id, finding_id = %finding_id))]
pub async fn get_impact( pub async fn get_impact(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path((repo_id, finding_id)): Path<(String, String)>, Path((repo_id, finding_id)): Path<(String, String)>,
@@ -183,7 +195,7 @@ pub async fn get_impact(
} }
/// GET /api/v1/graph/:repo_id/search — BM25 symbol search /// GET /api/v1/graph/:repo_id/search — BM25 symbol search
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %repo_id, query = %params.q))]
pub async fn search_symbols( pub async fn search_symbols(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(repo_id): Path<String>, Path(repo_id): Path<String>,
@@ -204,7 +216,10 @@ pub async fn search_symbols(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to search graph nodes: {e}");
Vec::new()
}
}; };
let total = nodes.len() as u64; let total = nodes.len() as u64;
@@ -216,7 +231,7 @@ pub async fn search_symbols(
} }
/// GET /api/v1/graph/:repo_id/file-content — Read source file from cloned repo /// GET /api/v1/graph/:repo_id/file-content — Read source file from cloned repo
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
pub async fn get_file_content( pub async fn get_file_content(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(repo_id): Path<String>, Path(repo_id): Path<String>,
@@ -278,7 +293,7 @@ pub struct FileContent {
} }
/// POST /api/v1/graph/:repo_id/build — Trigger graph rebuild /// POST /api/v1/graph/:repo_id/build — Trigger graph rebuild
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
pub async fn trigger_build( pub async fn trigger_build(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(repo_id): Path<String>, Path(repo_id): Path<String>,

View File

@@ -234,7 +234,10 @@ pub async fn stats_overview(Extension(agent): AgentExt) -> ApiResult<OverviewSta
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch recent scans: {e}");
Vec::new()
}
}; };
Ok(Json(ApiResponse { Ok(Json(ApiResponse {
@@ -276,7 +279,10 @@ pub async fn list_repositories(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch repositories: {e}");
Vec::new()
}
}; };
Ok(Json(ApiResponse { Ok(Json(ApiResponse {
@@ -342,7 +348,7 @@ pub async fn get_ssh_public_key(
Ok(Json(serde_json::json!({ "public_key": public_key.trim() }))) Ok(Json(serde_json::json!({ "public_key": public_key.trim() })))
} }
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %id))]
pub async fn trigger_scan( pub async fn trigger_scan(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(id): Path<String>, Path(id): Path<String>,
@@ -357,7 +363,7 @@ pub async fn trigger_scan(
Ok(Json(serde_json::json!({ "status": "scan_triggered" }))) Ok(Json(serde_json::json!({ "status": "scan_triggered" })))
} }
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %id))]
pub async fn delete_repository( pub async fn delete_repository(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(id): Path<String>, Path(id): Path<String>,
@@ -404,7 +410,7 @@ pub async fn delete_repository(
Ok(Json(serde_json::json!({ "status": "deleted" }))) Ok(Json(serde_json::json!({ "status": "deleted" })))
} }
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = ?filter.repo_id, severity = ?filter.severity, scan_type = ?filter.scan_type))]
pub async fn list_findings( pub async fn list_findings(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Query(filter): Query<FindingsFilter>, Query(filter): Query<FindingsFilter>,
@@ -463,7 +469,10 @@ pub async fn list_findings(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch findings: {e}");
Vec::new()
}
}; };
Ok(Json(ApiResponse { Ok(Json(ApiResponse {
@@ -473,7 +482,7 @@ pub async fn list_findings(
})) }))
} }
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(finding_id = %id))]
pub async fn get_finding( pub async fn get_finding(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(id): Path<String>, Path(id): Path<String>,
@@ -494,7 +503,7 @@ pub async fn get_finding(
})) }))
} }
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(finding_id = %id))]
pub async fn update_finding_status( pub async fn update_finding_status(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Path(id): Path<String>, Path(id): Path<String>,
@@ -598,7 +607,7 @@ pub async fn sbom_filters(
}))) })))
} }
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = ?filter.repo_id, package_manager = ?filter.package_manager))]
pub async fn list_sbom( pub async fn list_sbom(
Extension(agent): AgentExt, Extension(agent): AgentExt,
Query(filter): Query<SbomFilter>, Query(filter): Query<SbomFilter>,
@@ -644,7 +653,10 @@ pub async fn list_sbom(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch SBOM entries: {e}");
Vec::new()
}
}; };
Ok(Json(ApiResponse { Ok(Json(ApiResponse {
@@ -666,7 +678,10 @@ pub async fn export_sbom(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch SBOM entries for export: {e}");
Vec::new()
}
}; };
let body = if params.format == "spdx" { let body = if params.format == "spdx" {
@@ -799,7 +814,10 @@ pub async fn license_summary(
let entries: Vec<SbomEntry> = match db.sbom_entries().find(query).await { let entries: Vec<SbomEntry> = match db.sbom_entries().find(query).await {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch SBOM entries for license summary: {e}");
Vec::new()
}
}; };
let mut license_map: std::collections::HashMap<String, Vec<String>> = let mut license_map: std::collections::HashMap<String, Vec<String>> =
@@ -845,7 +863,10 @@ pub async fn sbom_diff(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch SBOM entries for repo_a: {e}");
Vec::new()
}
}; };
let entries_b: Vec<SbomEntry> = match db let entries_b: Vec<SbomEntry> = match db
@@ -854,7 +875,10 @@ pub async fn sbom_diff(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch SBOM entries for repo_b: {e}");
Vec::new()
}
}; };
// Build maps by (name, package_manager) -> version // Build maps by (name, package_manager) -> version
@@ -944,7 +968,10 @@ pub async fn list_issues(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch tracker issues: {e}");
Vec::new()
}
}; };
Ok(Json(ApiResponse { Ok(Json(ApiResponse {
@@ -972,7 +999,10 @@ pub async fn list_scan_runs(
.await .await
{ {
Ok(cursor) => collect_cursor_async(cursor).await, Ok(cursor) => collect_cursor_async(cursor).await,
Err(_) => Vec::new(), Err(e) => {
tracing::warn!("Failed to fetch scan runs: {e}");
Vec::new()
}
}; };
Ok(Json(ApiResponse { Ok(Json(ApiResponse {

View File

@@ -27,6 +27,7 @@ impl CveScanner {
repo_id: &str, repo_id: &str,
entries: &mut [SbomEntry], entries: &mut [SbomEntry],
) -> Result<Vec<CveAlert>, CoreError> { ) -> Result<Vec<CveAlert>, CoreError> {
tracing::info!("scanning {} SBOM entries for known CVEs", entries.len());
let mut alerts = Vec::new(); let mut alerts = Vec::new();
// Batch query OSV.dev // Batch query OSV.dev
@@ -93,7 +94,10 @@ impl CveScanner {
.json(&body) .json(&body)
.send() .send()
.await .await
.map_err(|e| CoreError::Http(format!("OSV.dev request failed: {e}")))?; .map_err(|e| {
tracing::warn!("OSV.dev API call failed: {e}");
CoreError::Http(format!("OSV.dev request failed: {e}"))
})?;
if !resp.status().is_success() { if !resp.status().is_success() {
let status = resp.status(); let status = resp.status();
@@ -104,10 +108,10 @@ impl CveScanner {
continue; continue;
} }
let result: OsvBatchResponse = resp let result: OsvBatchResponse = resp.json().await.map_err(|e| {
.json() tracing::warn!("failed to parse OSV.dev response: {e}");
.await CoreError::Http(format!("Failed to parse OSV.dev response: {e}"))
.map_err(|e| CoreError::Http(format!("Failed to parse OSV.dev response: {e}")))?; })?;
let chunk_vulns = result.results.into_iter().map(|r| { let chunk_vulns = result.results.into_iter().map(|r| {
r.vulns r.vulns

View File

@@ -78,10 +78,12 @@ 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> { pub fn clone_or_fetch(&self, git_url: &str, repo_name: &str) -> Result<PathBuf, AgentError> {
let repo_path = self.base_path.join(repo_name); let repo_path = self.base_path.join(repo_name);
if repo_path.exists() { if repo_path.exists() {
tracing::info!("fetching updates for existing repo");
self.fetch(&repo_path)?; self.fetch(&repo_path)?;
} else { } else {
std::fs::create_dir_all(&repo_path)?; std::fs::create_dir_all(&repo_path)?;
@@ -92,7 +94,9 @@ impl GitOps {
Ok(repo_path) Ok(repo_path)
} }
#[tracing::instrument(skip_all)]
fn clone_repo(&self, git_url: &str, repo_path: &Path) -> Result<(), AgentError> { fn clone_repo(&self, git_url: &str, repo_path: &Path) -> Result<(), AgentError> {
tracing::info!("cloning repo from {}", git_url);
let mut builder = git2::build::RepoBuilder::new(); let mut builder = git2::build::RepoBuilder::new();
let fetch_opts = self.credentials.fetch_options(); let fetch_opts = self.credentials.fetch_options();
builder.fetch_options(fetch_opts); builder.fetch_options(fetch_opts);

View File

@@ -1,6 +1,7 @@
use std::sync::Arc; use std::sync::Arc;
use mongodb::bson::doc; use mongodb::bson::doc;
use tracing::Instrument;
use compliance_core::models::*; use compliance_core::models::*;
use compliance_core::traits::Scanner; use compliance_core::traits::Scanner;
@@ -50,7 +51,7 @@ impl PipelineOrchestrator {
} }
} }
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = %repo_id, trigger = ?trigger))]
pub async fn run(&self, repo_id: &str, trigger: ScanTrigger) -> Result<(), AgentError> { pub async fn run(&self, repo_id: &str, trigger: ScanTrigger) -> Result<(), AgentError> {
// Look up the repository // Look up the repository
let repo = self let repo = self
@@ -90,6 +91,7 @@ impl PipelineOrchestrator {
.await?; .await?;
} }
Err(e) => { Err(e) => {
tracing::error!(repo_id, error = %e, "Scan pipeline failed");
self.db self.db
.scan_runs() .scan_runs()
.update_one( .update_one(
@@ -109,7 +111,7 @@ impl PipelineOrchestrator {
result.map(|_| ()) result.map(|_| ())
} }
#[tracing::instrument(skip_all)] #[tracing::instrument(skip_all, fields(repo_id = repo.name.as_str()))]
async fn run_pipeline( async fn run_pipeline(
&self, &self,
repo: &TrackedRepository, repo: &TrackedRepository,
@@ -138,8 +140,13 @@ impl PipelineOrchestrator {
// Stage 1: Semgrep SAST // Stage 1: Semgrep SAST
tracing::info!("[{repo_id}] Stage 1: Semgrep SAST"); tracing::info!("[{repo_id}] Stage 1: Semgrep SAST");
self.update_phase(scan_run_id, "sast").await; self.update_phase(scan_run_id, "sast").await;
let semgrep = SemgrepScanner; match async {
match semgrep.scan(&repo_path, &repo_id).await { let semgrep = SemgrepScanner;
semgrep.scan(&repo_path, &repo_id).await
}
.instrument(tracing::info_span!("stage_sast"))
.await
{
Ok(output) => all_findings.extend(output.findings), Ok(output) => all_findings.extend(output.findings),
Err(e) => tracing::warn!("[{repo_id}] Semgrep failed: {e}"), Err(e) => tracing::warn!("[{repo_id}] Semgrep failed: {e}"),
} }
@@ -147,8 +154,13 @@ impl PipelineOrchestrator {
// Stage 2: SBOM Generation // Stage 2: SBOM Generation
tracing::info!("[{repo_id}] Stage 2: SBOM Generation"); tracing::info!("[{repo_id}] Stage 2: SBOM Generation");
self.update_phase(scan_run_id, "sbom_generation").await; self.update_phase(scan_run_id, "sbom_generation").await;
let sbom_scanner = SbomScanner; let mut sbom_entries = match async {
let mut sbom_entries = match sbom_scanner.scan(&repo_path, &repo_id).await { let sbom_scanner = SbomScanner;
sbom_scanner.scan(&repo_path, &repo_id).await
}
.instrument(tracing::info_span!("stage_sbom_generation"))
.await
{
Ok(output) => output.sbom_entries, Ok(output) => output.sbom_entries,
Err(e) => { Err(e) => {
tracing::warn!("[{repo_id}] SBOM generation failed: {e}"); tracing::warn!("[{repo_id}] SBOM generation failed: {e}");
@@ -167,9 +179,13 @@ impl PipelineOrchestrator {
k.expose_secret().to_string() k.expose_secret().to_string()
}), }),
); );
let cve_alerts = match cve_scanner let cve_alerts = match async {
.scan_dependencies(&repo_id, &mut sbom_entries) cve_scanner
.await .scan_dependencies(&repo_id, &mut sbom_entries)
.await
}
.instrument(tracing::info_span!("stage_cve_scanning"))
.await
{ {
Ok(alerts) => alerts, Ok(alerts) => alerts,
Err(e) => { Err(e) => {
@@ -181,22 +197,36 @@ impl PipelineOrchestrator {
// Stage 4: Pattern Scanning (GDPR + OAuth) // Stage 4: Pattern Scanning (GDPR + OAuth)
tracing::info!("[{repo_id}] Stage 4: Pattern Scanning"); tracing::info!("[{repo_id}] Stage 4: Pattern Scanning");
self.update_phase(scan_run_id, "pattern_scanning").await; self.update_phase(scan_run_id, "pattern_scanning").await;
let gdpr = GdprPatternScanner::new(); {
match gdpr.scan(&repo_path, &repo_id).await { let pattern_findings = async {
Ok(output) => all_findings.extend(output.findings), let mut findings = Vec::new();
Err(e) => tracing::warn!("[{repo_id}] GDPR pattern scan failed: {e}"), let gdpr = GdprPatternScanner::new();
} match gdpr.scan(&repo_path, &repo_id).await {
let oauth = OAuthPatternScanner::new(); Ok(output) => findings.extend(output.findings),
match oauth.scan(&repo_path, &repo_id).await { Err(e) => tracing::warn!("[{repo_id}] GDPR pattern scan failed: {e}"),
Ok(output) => all_findings.extend(output.findings), }
Err(e) => tracing::warn!("[{repo_id}] OAuth pattern scan failed: {e}"), let oauth = OAuthPatternScanner::new();
match oauth.scan(&repo_path, &repo_id).await {
Ok(output) => findings.extend(output.findings),
Err(e) => tracing::warn!("[{repo_id}] OAuth pattern scan failed: {e}"),
}
findings
}
.instrument(tracing::info_span!("stage_pattern_scanning"))
.await;
all_findings.extend(pattern_findings);
} }
// Stage 4a: Secret Detection (Gitleaks) // Stage 4a: Secret Detection (Gitleaks)
tracing::info!("[{repo_id}] Stage 4a: Secret Detection"); tracing::info!("[{repo_id}] Stage 4a: Secret Detection");
self.update_phase(scan_run_id, "secret_detection").await; self.update_phase(scan_run_id, "secret_detection").await;
let gitleaks = GitleaksScanner; match async {
match gitleaks.scan(&repo_path, &repo_id).await { let gitleaks = GitleaksScanner;
gitleaks.scan(&repo_path, &repo_id).await
}
.instrument(tracing::info_span!("stage_secret_detection"))
.await
{
Ok(output) => all_findings.extend(output.findings), Ok(output) => all_findings.extend(output.findings),
Err(e) => tracing::warn!("[{repo_id}] Gitleaks failed: {e}"), Err(e) => tracing::warn!("[{repo_id}] Gitleaks failed: {e}"),
} }
@@ -204,8 +234,13 @@ impl PipelineOrchestrator {
// Stage 4b: Lint Scanning // Stage 4b: Lint Scanning
tracing::info!("[{repo_id}] Stage 4b: Lint Scanning"); tracing::info!("[{repo_id}] Stage 4b: Lint Scanning");
self.update_phase(scan_run_id, "lint_scanning").await; self.update_phase(scan_run_id, "lint_scanning").await;
let lint = LintScanner; match async {
match lint.scan(&repo_path, &repo_id).await { let lint = LintScanner;
lint.scan(&repo_path, &repo_id).await
}
.instrument(tracing::info_span!("stage_lint_scanning"))
.await
{
Ok(output) => all_findings.extend(output.findings), Ok(output) => all_findings.extend(output.findings),
Err(e) => tracing::warn!("[{repo_id}] Lint scanning failed: {e}"), Err(e) => tracing::warn!("[{repo_id}] Lint scanning failed: {e}"),
} }
@@ -214,19 +249,26 @@ impl PipelineOrchestrator {
if let Some(old_sha) = &repo.last_scanned_commit { if let Some(old_sha) = &repo.last_scanned_commit {
tracing::info!("[{repo_id}] Stage 4c: LLM Code Review"); tracing::info!("[{repo_id}] Stage 4c: LLM Code Review");
self.update_phase(scan_run_id, "code_review").await; self.update_phase(scan_run_id, "code_review").await;
let reviewer = CodeReviewScanner::new(self.llm.clone()); let review_output = async {
let review_output = reviewer let reviewer = CodeReviewScanner::new(self.llm.clone());
.review_diff(&repo_path, &repo_id, old_sha, &current_sha) reviewer
.await; .review_diff(&repo_path, &repo_id, old_sha, &current_sha)
.await
}
.instrument(tracing::info_span!("stage_code_review"))
.await;
all_findings.extend(review_output.findings); all_findings.extend(review_output.findings);
} }
// Stage 4.5: Graph Building // Stage 4.5: Graph Building
tracing::info!("[{repo_id}] Stage 4.5: Graph Building"); tracing::info!("[{repo_id}] Stage 4.5: Graph Building");
self.update_phase(scan_run_id, "graph_building").await; self.update_phase(scan_run_id, "graph_building").await;
let graph_context = match self let graph_context = match async {
.build_code_graph(&repo_path, &repo_id, &all_findings) self.build_code_graph(&repo_path, &repo_id, &all_findings)
.await .await
}
.instrument(tracing::info_span!("stage_graph_building"))
.await
{ {
Ok(ctx) => Some(ctx), Ok(ctx) => Some(ctx),
Err(e) => { Err(e) => {

View File

@@ -46,6 +46,7 @@ impl Scanner for SbomScanner {
/// Generate missing lock files so Syft can resolve the full dependency tree. /// Generate missing lock files so Syft can resolve the full dependency tree.
/// This handles repos that gitignore their lock files (common for Rust libraries). /// This handles repos that gitignore their lock files (common for Rust libraries).
#[tracing::instrument(skip_all)]
async fn generate_lockfiles(repo_path: &Path) { async fn generate_lockfiles(repo_path: &Path) {
// Cargo: generate Cargo.lock if Cargo.toml exists without it // Cargo: generate Cargo.lock if Cargo.toml exists without it
if repo_path.join("Cargo.toml").exists() && !repo_path.join("Cargo.lock").exists() { if repo_path.join("Cargo.toml").exists() && !repo_path.join("Cargo.lock").exists() {
@@ -122,6 +123,7 @@ async fn generate_lockfiles(repo_path: &Path) {
/// Enrich Cargo SBOM entries with license info from `cargo metadata`. /// Enrich Cargo SBOM entries with license info from `cargo metadata`.
/// Syft doesn't read license data from Cargo.lock, so we fill it in. /// Syft doesn't read license data from Cargo.lock, so we fill it in.
#[tracing::instrument(skip_all)]
async fn enrich_cargo_licenses(repo_path: &Path, entries: &mut [SbomEntry]) { async fn enrich_cargo_licenses(repo_path: &Path, entries: &mut [SbomEntry]) {
if !repo_path.join("Cargo.toml").exists() { if !repo_path.join("Cargo.toml").exists() {
return; return;
@@ -182,6 +184,7 @@ async fn enrich_cargo_licenses(repo_path: &Path, entries: &mut [SbomEntry]) {
} }
} }
#[tracing::instrument(skip_all, fields(repo_id = %repo_id))]
async fn run_syft(repo_path: &Path, repo_id: &str) -> Result<Vec<SbomEntry>, CoreError> { async fn run_syft(repo_path: &Path, repo_id: &str) -> Result<Vec<SbomEntry>, CoreError> {
let output = tokio::process::Command::new("syft") let output = tokio::process::Command::new("syft")
.arg(repo_path) .arg(repo_path)
@@ -232,6 +235,7 @@ async fn run_syft(repo_path: &Path, repo_id: &str) -> Result<Vec<SbomEntry>, Cor
Ok(entries) Ok(entries)
} }
#[tracing::instrument(skip_all)]
async fn run_cargo_audit(repo_path: &Path, _repo_id: &str) -> Result<Vec<AuditVuln>, CoreError> { async fn run_cargo_audit(repo_path: &Path, _repo_id: &str) -> Result<Vec<AuditVuln>, CoreError> {
let cargo_lock = repo_path.join("Cargo.lock"); let cargo_lock = repo_path.join("Cargo.lock");
if !cargo_lock.exists() { if !cargo_lock.exists() {