More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Context Audit: What's Eating Your Context

    1

    Measures what your coding agent loads before you type (rules files, skills, MCP servers) and shows the biggest token costs.

    $9

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    About this skill

    Measures how much of an AI coding agent's context window is used before the user types anything. It covers always-loaded instruction files for every host (CLAUDE.md with @imports and the memory index, AGENTS.md, GEMINI.md, always-apply Cursor rules, Copilot instructions), every installed skill's listed description, subagent descriptions and configured MCP servers.

    Use when your agent feels slow, forgetful or expensive, you want to know how big your CLAUDE.md or context is, what's eating your context window, or which skills or MCP servers you don't use. It only measures and reports; to remove or disable anything, use context-trim (in the full pack).

    Works in Claude Code, Codex, Gemini CLI, Cursor and Copilot. Live-tested on Claude Code, Codex and Gemini CLI (eval grade A). Part of Context Budget Audit by Abiodun Anifowose: the full pack with every related skill is at

    Requires Python 3. Exact token counting requires an ANTHROPIC_API_KEY. Usage analysis only reads local skill invocation names, not message content.

    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

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 25 days ago

    Requires

    Anthropic
    Python

    Permissions

    Terminal / Shell
    Read Files
    Write Files
    Network Access
    Environment Variables

    Allowed Hosts

    api.anthropic.com
    Listed25 days ago

    What's inside

    Frequently Asked Questions

    The Context Audit skill scans your configuration files, system prompts, and plugin descriptions to identify redundant instructions and 'dead weight' tokens that cause agents to become slow, forgetful, or expensive.

    It is compatible with major AI development environments including Claude Code, Cursor, GitHub Copilot, and Gemini CLI, as well as any agent framework that uses local configuration files like CLAUDE.md.

    While the tool provides close estimates offline, you will need a valid Anthropic API key if you want to perform 1:1 exact token counting for Claude-based models.

    No, the audit is non-destructive; it generates a detailed report (Markdown or HTML) highlighting where you can safely trim or merge files without automatically altering your project code.

    You receive the full Python-based audit tool, integration scripts for supported IDEs, and a set of performance templates to help you benchmark your agent's efficiency.

    The tool requires a local Python 3 environment and focuses on auditing your configuration and session logs; it does not require access to your live production databases or private source code logic.

    New in LLM Ops