agentleFS
desertcachedesertcache/agent-os/skills/audit-instructions/SKILL.md

audit-instructions

Skill0 starsChanged 9 months agoMITView on GitHub

Self-review: finds stale, missing, redundant, or conflicting rules across all config files. Use when the user says "audit instructions", "check rules", or "/audit-instructions". Reads CLAUDE.md, MEMORY.md, error-patterns.md, BRAIN.md, all skill files, and settings.json to produce a categorized report with suggested fixes.

2 other projects use a version of this.

Adopted by

Most recently taken up by

  • Other versions2 projects use a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 2

    1. kartykk/agent-ready★ 0same name
    2. kitloomlk-stack/agent-skill-builder-kit★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository9 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

10 rules. About 0 are common, 8 are distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read ALL of these files (use parallel reads where possible).

    line 16
  2. Very specific:

    All skill files: ~/.claude/skills/*/SKILL.md

    line 24
  3. Distinctive:

    Do codebase paths referenced in CLAUDE.md and skill files still exist on disk?

    line 30
  4. Distinctive:

    Do all skills reference files that actually exist?

    line 36
  5. Distinctive:

    Do any files give conflicting paths or project names?

    line 47
  6. Distinctive:

    After presenting the report, ask: "Want me to apply the suggested fixes?"

    line 75
  7. Distinctive:

    Read files first, report second — don't guess about file contents

    line 78
  8. Distinctive:

    Don't auto-fix anything — always ask first

    line 81
  9. Distinctive:

    If everything looks clean, say so: "All instructions consistent.

    line 82
  10. Distinctive:

    Keep the report concise — group similar issues rather than listing each individually

    line 83

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Audit Instructions

Perform a thorough self-review of all configuration and instruction files. Produce an actionable report. Do NOT apply fixes without user confirmation.

Files to Read

Read ALL of these files (use parallel reads where possible). Paths are resolved from agent-os.config.json:

  1. ~/.claude/CLAUDE.md — main instruction file
  2. <memoryDir>/MEMORY.md — auto memory
  3. <memoryDir>/error-patterns.md — error learnings
  4. <memoryDir>/patterns.md — workflow patterns
  5. <memoryDir>/debugging.md — debugging strategies
  6. <vaultPath>/BRAIN.md — vault brain file
  7. All skill files: ~/.claude/skills/*/SKILL.md
  8. ~/.claude/settings.json — hooks and settings

Checks to Perform

STALE — Information that may be outdated

  • Do codebase paths referenced in CLAUDE.md and skill files still exist on disk? (Use Glob to verify key directories)
  • Is the project list current? Check if any projects in vault Projects/ are missing from the mapping table
  • Is MEMORY.md approaching the 200-line limit? (Count lines, warn if >180)
  • Are there session handoff notes >30 days old that could be archived?

MISSING — Things that should exist but don't

  • Do all skills reference files that actually exist? (e.g., cache files, vault paths)
  • Are there common error patterns from recent sessions not captured in error-patterns.md?
  • Are there projects in the codebase mapping that lack a vault folder?
  • Are there hook scripts referenced in settings.json that don't exist on disk?

REDUNDANT — Same info in too many places

  • Project-codebase mapping: canonical source is CLAUDE.md. The session-context-preloader hook has its own copy (in agent-os.config.json). Flag if they're out of sync.
  • Are there duplicate rules or overlapping instructions between CLAUDE.md and MEMORY.md?
  • Are there error patterns that are also mentioned in MEMORY.md (redundant)?

CONFLICT — Contradictory information

  • Do any files give conflicting paths or project names?
  • Are there behavioral instructions that contradict each other?

Output Format

Instruction Audit Report
========================

STALE (N issues)
- [file:line] Description of what's stale
  Suggested fix: ...

MISSING (N issues)
- [file] Description of what's missing
  Suggested fix: ...

REDUNDANT (N issues)
- [file1 + file2] Description of redundancy
  Suggested fix: ...

CONFLICT (N issues)
- [file1 vs file2] Description of conflict
  Suggested fix: ...

Summary: N total issues (N critical, N minor)

After presenting the report, ask: "Want me to apply the suggested fixes?"

Rules

  • Read files first, report second — don't guess about file contents
  • Be specific: include file paths and line numbers where possible
  • Categorize severity: CRITICAL (will cause errors), MINOR (cosmetic/organizational)
  • Don't auto-fix anything — always ask first
  • If everything looks clean, say so: "All instructions consistent. No issues found."
  • Keep the report concise — group similar issues rather than listing each individually

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.