PR and Baseline Workflows

Audit only the PR diff:

maintainability-agent \
  --changed-only main...HEAD \
  --output maintainability-report.md \
  --prompt-output maintainability-remediation-prompt.md \
  --comment-output maintainability-pr-comment.md

Create a baseline for existing debt:

maintainability-agent \
  --write-baseline maintainability-baseline.json

Fail only on findings not present in the baseline:

maintainability-agent \
  --baseline maintainability-baseline.json \
  --fail-on-new

Generate reusable AI coding-agent instructions:

maintainability-agent \
  --agent-instructions-output AGENTS-maintainability.md

Generate persistent agent standards before code is written:

maintainability-audit \
  --config maintainability-agent.json \
  --init-agent-standards \
  --target codex \
  --target claude-code \
  --target cursor \
  --target copilot \
  --target windsurf \
  --instructions-output-dir .

This writes tool-native instruction files such as AGENTS.md, CLAUDE.md, .cursor/rules/maintainability.mdc, .github/copilot-instructions.md, and .windsurf/rules/maintainability.md.

Check a remediating diff against the work order that produced it:

maintainability-agent \
  --conformance main...HEAD \
  --fail-on-out-of-scope \
  --fail-on-regression \
  --attestation-output maintainability-attestation.md

clean fails on a suppression directive the diff added to a file the work order named, and — since 3.11.0 — on any test file in the diff made unable to fail: a skip or suppression, an assertion that cannot fail, the assertion API replaced, a failing assertion swallowed, or assertions deleted with nothing in their place (_oracle, reported as oracle_weakened). Held to a precision bar frozen before the detector was written. For work this tool did not order, pass the task's ask with --ask FILE.