Add compliance_report stage emitting 4 per-standard markdown reports (Phase 3'.1 B-4)#11
Merged
Merged
Conversation
…(Phase 3'.1 B-4)
Closes the second-to-last Phase 3'.1 step. The new stage takes the
four-document compatibility-mapping suite (B-1..B-3) and turns it
into a per-run deliverable: every analysis run now emits four
standard-specific markdown reports under
stages/compliance_report/, alongside a structured stage.json that
records the evidence sources consulted and the aggregate counts.
- src/aiedge/compliance_report.py (new, 43rd pipeline stage)
* ComplianceEvidence dataclass aggregates per-stage counts:
sbom_component_count, cve_match_count, cve_critical_high_count,
finding_total, finding_by_{severity,category,evidence_tier},
cert_finding_count, init_high_risk_service_count,
fs_perm_finding_count, evidence sources dict.
* _gather_evidence() reads sbom.json, cve_matches.json,
findings.json, certificate_analysis.json, init_analysis.json,
fs_permissions.json. Every input is optional; missing
artefacts yield zero counts.
* _render_report() produces a self-contained markdown with the
mandatory "compatible with" disclaimer, evidence summary,
severity / category / evidence-tier tables, and a pointer back
to the canonical mapping document under
docs/compliance_mapping/.
* Stage degrades to status=partial (without crashing) when no
upstream artefacts are present, so the four reports always
land.
- src/aiedge/stage_registry.py
* _make_compliance_report_stage() factory registered as
"compliance_report" in _STAGE_FACTORIES (43 entries total).
- src/aiedge/stage_dag.py
* STAGE_DEPS adds "compliance_report" depending on
exploit_policy + sbom + cve_scan so the stage always runs
after the major evidence-producing stages.
- tests/test_compliance_report.py (new, 8 tests)
* Stage emits 4 reports; degrades to partial without evidence;
evidence counts match seeded inputs; sources only list
present artefacts; every report carries the "compatible with"
disclaimer; every report links back to the canonical mapping;
stage_registry exposes the factory; factory returns the
expected stage.
Verification:
pytest -q tests/test_compliance_report.py # 8 passed
pytest -q # full suite green (incl. test_stage_dag)
ruff check, pyright (changed files) # clean / 0 errors
python3 scripts/check_doc_consistency.py # OK
Phase 3'.1 Track B fourth step (Pivot 2026-04-19).
B-5 (v2.7.1 release tag) remains as the closing step after Track A
v2.7.0 lands.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
…port-stage # Conflicts: # CHANGELOG.md
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Phase 3'.1 Track B fourth PR. Adds a new 43rd pipeline stage that turns the four-document compatibility-mapping suite (PR #10: B-1..B-3) into a per-run deliverable.
Every analysis run now emits four standard-specific markdown reports:
Each report aggregates per-run evidence counts from
sbom,cve_scan,findings,cert_analysis,init_analysis, andfs_permissions, then links back to the canonical mapping document underdocs/compliance_mapping/.Mandatory wording
Every emitted report carries the
"compatible with"disclaimer required byscripts/check_doc_consistency.pyrule 3. Tests assert both presence of the canonical phrase and absence of the forbidden"compliant with"/"fully compliant"substitutions.Stage registration
_make_compliance_report_stage()factory registered as"compliance_report"in_STAGE_FACTORIES(43 entries).STAGE_DEPS["compliance_report"] = frozenset({"exploit_policy", "sbom", "cve_scan"})so the stage always runs after the major evidence-producing stages.Degradation behaviour
When no upstream artefacts are present (e.g. an
--no-llmsmoke run with most stages skipped), the stage degrades tostatus=partial(without crashing) and still emits all four reports with zero counts plus an explicit limitation. This guarantees the four files always land.Test plan
pytest -q tests/test_compliance_report.py— 8 passed (stage emits 4 reports / degrades to partial without evidence / evidence counts match seeded inputs / sources only list present artefacts / every report carries the "compatible with" disclaimer / every report links back to the canonical mapping / stage_registry exposes the factory / factory returns the expected stage)pytest -q— full suite green (includingtest_stage_dag::test_stage_deps_keys_match_factories_exact)ruff check src/aiedge/compliance_report.py src/aiedge/stage_registry.py src/aiedge/stage_dag.py tests/test_compliance_report.py— cleanpyright— 0 errors on changed filespython3 scripts/check_doc_consistency.py— OKFiles
src/aiedge/compliance_report.py(new, 380 lines)src/aiedge/stage_registry.py(+12 lines: factory + dict entry)src/aiedge/stage_dag.py(+1 line: STAGE_DEPS entry)tests/test_compliance_report.py(new, 280 lines)CHANGELOG.md(Unreleased ### Added)Why a separate PR from PR #10
PR #10 was docs-only (the four mapping markdown files). This PR is code-only (stage + factory + DAG + tests). Splitting keeps each PR's review scope focused and lets the docs PR merge independently of any code-review delay on the stage.
Plan reference
See
~/.claude/plans/twinkly-hugging-leaf.md(Track B B-4) for the broader Track A / Track B parallel plan.🤖 Generated with Claude Code