Skip to content

GOVERNANCE_ARTIFACTS_README.md

Latest commit

 

History

History
120 lines (84 loc) · 3.45 KB

GOVERNANCE_ARTIFACTS_README.md

File metadata and controls

120 lines (84 loc) · 3.45 KB

Governance Artifacts Quickstart

This repository includes a daily G-SIFI AGI/ASI governance package with canonical artifacts, policy sketches, and validation tooling.

Files

  • DAILY_GSIFI_AGI_ASI_GOVERNANCE_2026_2030.md
  • artifacts/daily_governance_report.schema.json
  • artifacts/daily_governance_report.example.json
  • policies/sentinel_governance.rego
  • tools/validate_governance_artifacts.py
  • test_governance_snippets.py
  • test_validate_governance_artifacts.py

Local Validation

Install governance check dependencies:

pip install -r requirements-governance-checks.txt

Run Python syntax checks:

make daily-gsifi-governance-pycompile

Run the full combined check suite:

make daily-gsifi-governance-checks

Run the canonical validator:

python tools/validate_governance_artifacts.py
# or
make daily-gsifi-governance-validate

Run the unit tests:

pytest -q test_governance_snippets.py test_validate_governance_artifacts.py test_run_gsifi_governance_checks.py test_generate_gsifi_governance_report.py test_daily_gsifi_governance_workflow.py
# or
make daily-gsifi-governance-test

Generate a JUnit report locally:

mkdir -p artifacts/test-results
pytest -q --junitxml=artifacts/test-results/gsifi-governance-tests.xml \
  test_governance_snippets.py test_validate_governance_artifacts.py test_run_gsifi_governance_checks.py test_generate_gsifi_governance_report.py test_daily_gsifi_governance_workflow.py

CI

CI workflow:

  • .github/workflows/daily-gsifi-governance-validation.yml

It runs:

  1. make daily-gsifi-governance-ci
  2. Uploads JUnit + JSON run-summary artifacts.

Generate JUnit + JSON run evidence:

make daily-gsifi-governance-evidence

This repository includes machine-readable governance artifacts for AI controls:

  • schemas/bbom.schema.json
  • schemas/arre_record.schema.json
  • artifacts/bbom/*.json
  • examples/arre/*.json
  • tools/validate_ai_governance_artifacts.py

Local validation

python -m pip install -r requirements-governance.txt
python tools/validate_ai_governance_artifacts.py

Custom paths

python tools/validate_ai_governance_artifacts.py \
  --bbom-dir artifacts/bbom \
  --arre-dir examples/arre \
  --arre-dir evidence/arre

CI

Validation is enforced in .github/workflows/governance-artifacts.yml.

The validator enforces both JSON Schema compliance and semantic checks (for example BBOM threshold conformance and ARRE period consistency).

Additional semantic checks include date format validation (via JSON Schema format checking) and duplicate ARRE evidence hash detection.

Generate an auditable machine-readable summary report:

python tools/validate_ai_governance_artifacts.py --report-file .reports/governance-validation.json

Report output includes discovered/checked counters, passed_files, failed_files, and errors for audit trails.

The GitHub Actions workflow also uploads the report as a CI artifact (governance-validation-report) for audit retention.

When schema loading fails, reports include fatal_error: "schema_load_failure" and the associated error in errors.

Report output also includes validator_version and overall status (passed/failed) for easier pipeline gating.

Reports include exit_code (0 for pass, 2 for validation failure) to simplify CI/CD policy gating.

Per-domain failure counters (bbom_failed, arre_failed) are included to support targeted remediation dashboards.