ToolRank
About
Score and optimize MCP tool definitions for AI agent discovery. Analyzes Findability, Clarity, Precision, and Efficiency.
Details
- Author
- imhiroki
- Categories
- Developer Tools, AI
Jump to
Setup
Install ToolRank in your MCP client (Claude Desktop, Cursor, Windsurf, and others).
Repository: https://github.com/imhiroki/toolrank
Follow the installation instructions in the repository README, then restart your MCP client.
Score, optimize, and monitor how AI agents discover and select your MCP tools.
Score Your Tools →·Framework·Ranking·Blog
We scanned 4,162 MCP servers. Here's what we found.
73% of MCP servers are invisible to AI agents.They have no tool definitions, no descriptions, no schema. When an agent searches for tools, these servers don't exist.
Sources:arXiv 2602.14878,arXiv 2602.18914
ATO (Agent Tool Optimization)is to the agent economy what SEO was to the search economy.
LLMO is Stage 1 of ATO — necessary but not sufficient.
toolrank.dev/score— paste your tool JSON or enter your Smithery server name.
from toolrank_score import score_server, format_report result = score_server("my-server", tools) print(format_report(result))
- "name": "get", - "description": "gets data from the api" + "name": "search_repositories", + "description": "Searches for GitHub repositories matching a query. + Useful for finding open-source projects or checking if a repo exists. + Returns name, description, stars, language, and URL.", + "inputSchema": { + "type": "object", + "properties": { + "query": { "type": "string", "description": "Search query" }, + "sort": { "type": "string", "enum": ["stars", "forks", "updated"] } + }, + "required": ["query"] + }
Score: 52 → 96.Five minutes of work. 3.6x selection advantage.
toolrank/ ├── packages/ │ ├── scoring/ # Level A engine (Python, zero-cost) │ │ ├── toolrank_score.py # 14 checks across 4 dimensions │ │ ├── level_c_score.py # Claude AI scoring (Pro) │ │ └── weights.json # Auto-calibrated weights │ ├── scanner/ # Ecosystem scanner │ │ ├── scanner_v3.py # Weekly full / daily diff │ │ ├── calibrate.py # Weight auto-adjustment │ │ └── auto_blog.py # Daily article generation │ ├── web/ # Astro site (toolrank.dev) │ ├── mcp-server/ # ToolRank MCP Server │ └── badge-worker/ # Dynamic badge SVG (CF Workers) └── .github/workflows/ # Automated pipelines
ToolRank is open source. The scoring logic is fully transparent and auditable.
- Report issues:GitHub Issues
- Scoring methodology:packages/scoring/toolrank_score.py
- Governance:GOVERNANCE.md·CHANGELOG.md
- ATO Framework:toolrank.dev/framework
⭐Star this repoif you find ToolRank useful — it helps others discover it.
If SEO is about being found by search engines, ATO is about being used by AI agents.
This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.
An MCP server to help AI assistants to answer questions and generate AccelByte Extend SDK code more effectively .
Instead of direct calling MCP tools, mcpcode server transforms MCP tool calls into TypeScript programs, enabling smarter, lower-latency orchestration by LLMs.
Supercharge your Agent with Semantic Code Intelligence and save 💰 in the process!
AI-to-AI code review platform — Claude, Codex, and Gemini cross-check each other via MCP, REST API, and CLI for consensus-based results.
A code sandbox for AI assistants to safely execute arbitrary code. Requires a 302AI API key for authentication.
Agent-native developer Q&A API with MCP + A2A endpoints for citations, job pickup, and answer submission.
MCP server to dynamically load Claude Code skills into AI agents
MCP bridge that lets Claude Code delegate heavy tasks to the Antigravity CLI (agy) — purpose-built tools, model routing with fallback, session continuity, and output truncation to save Claude's context and tokens.
Exposes local execution of Aider commands as a detached process using MCP.
An MCP server for offloading AI coding tasks to Aider, enhancing development efficiency and flexibility.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
