Audit Fleet Evidence Policy

Ensure every finding is verifiable, traceable, and decision-ready.

Canonical Contract Terms

All audit-fleet outputs must preserve this exact section order:

1. Executive Summary
2. Findings
3. Quick Wins
4. High-Impact Expansions

All Findings entries must include these keys exactly:

• finding_id
• severity
• dimension
• evidence
• impact
• recommendation
• effort
• owner
• dependencies
• confidence
• acceptance_criteria

Severity enum is strict: critical | warning | info.

Evidence Standard

• evidence must include file:line plus a short quote or command output.
• impact must describe concrete consequence for delivery, security, UX, reliability, or business value.
• recommendation must define a single actionable next step.
• owner and dependencies must make execution ownership explicit.
• acceptance_criteria must define objective completion signal.

Confidence Guidance

• high: multiple corroborating evidence points.
• medium: one strong evidence point with limited ambiguity.
• low: weak or incomplete evidence.

Rejection Criteria

Reject any Findings row that misses required keys or uses severity outside critical | warning | info.