Instruction Slimmer

    1

    Rewrites CLAUDE.md and other instruction files to be shorter with zero behavior loss.

    Free

    Security scanned

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Instruction Slimmer

    Example session with this skill installed

    My CLAUDE.md has grown huge over time. /slim-instructions — tighten it without changing behavior.

    • Read your context and instructions
    • Compiled the instruction slimmer

    CLAUDE.md — 2,400 tokens → 980 tokens (59% reduction)

    What was cut:

    • 3 rules restated Claude's default behavior (removed — Claude already does these)
    • "You should make sure to always..." pattern simplified to "Always..." throughout
    • Two nearly-identical examples collapsed to one
    • "Note that:" and "Important:" meta-prefixes removed (added no information)

    Condensed version:

    
    # Project guidelines
    
    Always use TypeScript strict mode. Never commit directly to main.
    Run `npm test` before every PR. Prefer composition over inheritance.
    

    💾 Total savings: ~1,420 tokens recovered across 1 file. Copy the rewrite above to apply it.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Reads all loaded instruction files (CLAUDE.md, project settings, skill definitions), identifies redundancy and verbosity, and produces a condensed rewrite of each one — preserving every behavior while cutting token cost. Shows exactly what was cut and why.

    When to use

    • Your CLAUDE.md has grown over time through repeated edits
    • You're hitting context limits and instructions are a significant share
    • You want to audit whether your instructions are actually adding value

    How it works

    Invoke it by name in your agent (in Claude Code: /slim-instructions). It works only from instruction files already loaded in the session and outputs the rewrite for you to review and paste — it doesn't edit your files, make network calls, or use external services.

    Part of Claude Token Inspector, an open-source (MIT) suite of 8 context-window skills by the original author. Source: https://github.com/msishariful/claude-token-inspector

    Token counts are estimates. Requires existing CLAUDE.md or .md instruction files in context to function.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Free forever

    No account required to browse

    Trust & safety

    Security scan

    Security scan passed 23 days ago

    • Free with an Agensi account

    Permissions

    Read Files
    Listed23 days ago

    What's inside

    Frequently Asked Questions

    By stripping meta-commentary and redundant phrasing, this skill reduces the number of tokens the LLM must process in every turn, leading to faster initial response times and lower API costs without changing how the agent performs tasks.

    It's a standard SKILL.md skill, built and tested on Claude Code with CLAUDE.md. It also works in other SKILL.md-compatible agents (Cursor, Codex CLI, Gemini CLI, etc.) on whatever instruction files they have loaded.

    It's instructed to keep every rule, file path, command and tool name, and it lists exactly what it cut and why so you can check. It never overwrites your files; you review the rewrite before applying it.

    The SKILL.md instructions file. It's free and open source (MIT); the same skill is part of the Claude Token Inspector repo on GitHub.

    No setup beyond installing the skill. It's user-invoked: run it by name (in Claude Code: /slim-instructions) and it analyzes the instruction files already loaded in your session.

    New in Prompt & Skill Engineering