What problem does it solve?
Rule Audit helps prevent confusing or unsafe AI coding outcomes by finding contradictions, redundancy, gaps, drift, deny mismatches, broken links, and staleness across your host adapter rules—especially after you add rules or run sync-rules.
Core Features & Use Cases
- Cross-rule contradiction detection: Identifies incompatible directives in the same context and recommends canonical precedence.
- Redundancy and gap auditing: Detects repeated requirements and coverage gaps using evidence from memory and rule sections.
- Host drift and integrity checks: Verifies mandatory rules are referenced by the active host instruction file and that related-rules targets resolve.
- Evidence-gated decision output: Produces bounded findings with confidence, blockers, and nextAction, returning BLOCKED when required proofs are missing.
Quick Start
Ask your AI to audit the current ruleset by providing the rule-audit trigger, the selected rule files, the active host instruction file, and the required proof/receipts, then request an evidence-backed BLOCKED/PASS/FAIL/PARTIAL decision from rule-audit.