agentleFS
Sign inStart free
yltw27yltw27/agent-hygiene/skills/agent-hygiene/SKILL.md

agent-hygiene

Skill1 starsChanged 6 months agoMITView on GitHub

Explicit-only audit of local coding-agent setup, installed skills, MCP servers, plugins, and recent usage signals. Use only when the user explicitly invokes $agent-hygiene.

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. Arcanea-Labs/Starlight-Intelligence-System★ 10same name
    2. frankxai/Starlight-Intelligence-System★ 10same 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.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Do not edit, disable, unlink, uninstall, or remove anything during the audit unless the user explicitly asks for a follow-up cleanup step.

    line 12
  2. Distinctive:

    Treat usage data as a signal, not proof.

    line 28
  3. Distinctive:

    Ask a clarifying question only if the user wants a non-default scope.

    line 32
  4. Distinctive:

    Run the bundled audit script from the skill directory:

    line 33
  5. Distinctive:

    Prefer pruning global auto-triggerable skills before plugin or system skills.

    line 53

≈ “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

Agent Hygiene

Invocation Policy

This skill must stay explicit-only. Keep agents/openai.yaml with policy.allow_implicit_invocation: false so the skill is not injected into normal sessions by default.

Do not edit, disable, unlink, uninstall, or remove anything during the audit unless the user explicitly asks for a follow-up cleanup step.

Scope

Audit the user's local coding-agent setup and recommend how to reduce context noise and tool-selection ambiguity.

Focus on:

  • Installed skills across Codex and common shared skill directories
  • Codex plugins and MCP servers
  • Duplicate or overlapping skills
  • Skills installed globally that should be project-local
  • Broad mandatory skills that may trigger too often
  • Explicit skill mentions and tool-call signals in recent Codex sessions
  • Skills with no visible recent use

Treat usage data as a signal, not proof. Implicit skill triggering may not be visible in logs, and old sessions may be compacted or absent.

Workflow

  1. Ask a clarifying question only if the user wants a non-default scope. Default to the last 30 days and the current machine.
  2. Run the bundled audit script from the skill directory:
python3 /path/to/agent-hygiene/scripts/audit_agent_hygiene.py --days 30
  1. Inspect the script output. Read specific SKILL.md files only when needed to verify broad descriptions, overlap, or agent-specific instructions.
  2. Produce recommendations. Do not run uninstall commands.

Recommendation Rubric

Use these categories:

  • Keep: recently used, narrow trigger, high-value workflow, or installed as part of an enabled plugin the user actively uses.
  • Move project-local: useful only in one repo or team context.
  • Make explicit-only: rare, broad, expensive, or potentially disruptive but still useful on demand.
  • Rewrite: useful skill with vague, overlapping, or "always/mandatory" trigger wording.
  • Remove: duplicate, stale, wrong-agent install, broken path, or no credible recent use.
  • Investigate: usage evidence is too thin to recommend confidently.

Prefer pruning global auto-triggerable skills before plugin or system skills. For plugin-provided skills, recommend disabling/removing the plugin only when the whole plugin appears unused.

Output

Return a concise audit:

  • Summary: counts and main risk.
  • Recommendations: grouped by action, with evidence.
  • Notes: limitations of the usage evidence.
  • Next command: one safe command to inspect or verify, not a destructive cleanup command.

Be direct. The goal is a smaller, higher-signal agent setup, not a complete inventory dump.

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.

Check your setup, free