Audits an existing coding-agent harness (CLAUDE.md, AGENTS.md, rules files, hooks, MCP and tool config, skills) against the project's past Claude Code sessions, then returns a ranked findings report plus proposed diffs applied only after…
npx skills add arm-pz/harness-audit
4 files and 454 lines, free under MIT. Built from arm-pz/harness-audit at 95b9809.
Quality record
Version, date and limits are read from the repo on every sync, and the check runs every night. Real-world runs are listed only with their evidence.
Audits an existing coding-agent harness (CLAUDE.md, AGENTS.md, rules files, hooks, MCP and tool config, skills) against the project's past Claude Code sessions, then returns a ranked findings report plus proposed diffs applied only after approval. Flags bloated instruction files, rules with no evidence they help, prose rules nothing enforces, unused or overlapping tools, stale docs, and corrections that recur across sessions without a rule.
Measured from the packed archive at commit 95b9809. A skill is plain text your agent reads, so this list is all of it.
| File | Lines | Bytes |
|---|---|---|
| SKILL.md | 87 | 5,895 B |
| references/checks.md | 59 | 3,974 B |
| scripts/inventory.py | 147 | 5,596 B |
| scripts/mine_sessions.py | 161 | 6,413 B |
| 4 files | 454 | 8 KB |
Read it before you install it: the skill lives at skills/harness-audit in arm-pz/harness-audit.