toolmetry

sujugithub/toolmetry

★ 0 stars TypeScript AI/LLM Updated today

Measure how well AI agents use your MCP server's tools — then rewrite the descriptions and prove the improvement with before/after data

View on GitHub → 🔍 Audit Wallet Slippage →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "toolmetry": {
      "command": "npx",
      "args": [
        "-y",
        "sujugithub/toolmetry"
      ]
    }
  }
}

README Excerpt

> Measure how well AI agents actually use your MCP server's tools — then automatically rewrite the tool descriptions and **prove** the improvement with before/after data. Agents choose tools using nothing but your tool **names, descriptions, and schemas**. When descriptions overlap, overpromise, or under-specify, agents pick the wrong tool, pass the wrong arguments, or pad every task with wasteful extra calls — and your server gets blamed for it.

toolmetry

Quick Install

README Excerpt

Related AI/LLM Servers