# Agents Md Optimizer

> Audit bloated agent-instruction files (CLAUDE.md, AGENTS.md, and their local/user-level variants) and rewrite them lean, or author a new one from scratch following the 200-line "recipe book" principle. Use this skill whenever the user mentions CLAUDE.md, AGENTS.md, agent instruction file, project memory file, context file, token budget, context bloat, "my CLAUDE.md is too big", trimming CLAUDE.md, auditing AGENTS.md, writing AGENTS.md for a new project, or any discussion of what belongs in the agent-instruction-file hierarchy (enterprise, project, user-level, project-local). Also trigger when such a file is shown in context and is visibly long (over 200 lines) or contains anti-patterns like embedded API docs, negative "Don't do X" rules, step-by-step tutorials, or preemptive file-preload instructions — even if the user doesn't explicitly ask for optimization, proactively offer to use it.

## Facts
- Page: https://tashan.sh/capability/skill-msapps-mobile-agents-md-optimizer
- tashan id: skill:MSApps-Mobile/agents-md-optimizer
- Source: https://github.com/MSApps-Mobile/claude-plugins
- Type: skill
- Category: security
- tashan score: not scored (catalogued only — too little public evidence)
- Adoption: 9.0
- Upkeep: 95.0
- Freshness: 89.0
- Evidence coverage: 84% of the inputs this score can use
- Health: active
- Instruction depth: not yet graded
- License: MIT
- Official: no

## Install

```sh
cp -r agents-md-optimizer ~/.claude/skills/
```

## Security audit
Not scanned. We audit npm-published capabilities; this one has no npm package we can resolve, or has not reached the queue. This is not a clean bill of health.

---
Measured 2026-08-21 by tashan (https://tashan.sh) from public evidence. Scorer s5.