Pattern Proposal: Evidence-Grounded Compliance Scoring

Summary

Computes structured compliance findings and health scores by verifying architectural deltas against explicit rule contexts.

Candidate Pattern

• ID: evidence-grounded-compliance-scoring
• Classification: intelligence
• Maturity: established
• Type: implemented

Observed Signals

No specific signals listed.

Evidence

Supporting Evidence

• src/cas_runner/governance/compliance_engine.py (ComplianceEngine) - Evaluates rules and produces metrics
• src/cas_runner/governance/models.py (ComplianceFinding, ComplianceMetrics, GovernanceHealth) - Typed models tracking findings and overall health
• src/cas_runner/dashboard/health_model.py (calculate_health, HealthStatus) - Calculates health categorization from metrics

Scope Assessment

• Evidence Strength: high
• Confidence: high
• Applies to: ideas-to-life-cas

Rationale

• Intent: Provide quantitative, traceable governance metrics rather than ungrounded qualitative LLM judgments.
• Context: Evaluated during pull request validations and automated governance audits.
• Force: Binary pass/fail checks do not give actionable feedback to agent workflows
• Force: Unstructured LLM reviews are difficult to track over time or display in monitoring dashboards

Traceability

Artifact source: proposed-patterns