{
  "markdown": "# AgentRank MCP server\n\n**MCP server for AI agents to search real, source-backed businesses.** Every result\ncarries a source URL and a trust score, so your agent acts on facts — never invented\nphone numbers or addresses. Coverage: **Brazil, US, Europe** (~2.7M businesses, growing).\n\nWorks with any [MCP](https://modelcontextprotocol.io) client (Claude Desktop, Cursor, …).\nPublic beta · free key, no invite · website: https://agentsafepath.com ·\nofficial MCP registry: `io.github.saviodamato/agentrank-mcp`.\n\n## Use cases\n- Business search / company lookup with source-backed data.\n- Local business search for agents (city + category + contact).\n- Company profile: contacts, address, website, provenance.\n- Source-backed company data instead of hallucinated details.\n\n## Tools\n\n- **`search_businesses(intent, limit=10)`** — find businesses by a natural-language\n  intent (e.g. `\"farmácia em Copacabana com telefone\"`, `\"software company in California\"`).\n  Returns ranked results with `fit_score`, `trust_score`, justification, risks and the\n  **source URL** of each fact.\n- **`get_business_profile(company_id)`** — full profile of a result: location, contact\n  channels, hours, offers, policies and evidence, each with provenance and source.\n\n## 1. Get an API key\n\nRegister your agent (free, no invite) at\n**https://agentsafepath.com/registrar.html** — you receive a key once. It identifies\nyour agent and is sent as `Authorization: Bearer <key>`.\n\n## 2. Add it to your MCP client\n\n### Claude Desktop\n\nEdit `claude_desktop_config.json` (Settings → Developer → Edit Config) and add:\n\n```json\n{\n  \"mcpServers\": {\n    \"agentrank\": {\n      \"command\": \"uvx\",\n      \"args\": [\"agentrank-mcp\"],\n      \"env\": { \"AGENTRANK_API_KEY\": \"ak_your_key_here\" }\n    }\n  }\n}\n```\n\nRestart Claude Desktop. You'll see `search_businesses` and `get_business_profile`\navailable. (`uvx` comes with [uv](https://docs.astral.sh/uv/); or use the \"From source\"\noption below.)\n\n### Cursor\n\nAdd the same block under `mcpServers` in your Cursor MCP settings.\n\n### From source (no packaging)\n\n```bash\npip install -r requirements.txt\nAGENTRANK_API_KEY=ak_your_key_here python -m agentrank_mcp.server\n```\n\nAnd point your client's `command` at `python` with `args`\n`[\"-m\", \"agentrank_mcp.server\"]` and the folder on `PYTHONPATH`.\n\n## Environment variables\n\n| Var | Required | Default |\n|---|---|---|\n| `AGENTRANK_API_KEY` | yes | — |\n| `AGENTRANK_BASE_URL` | no | `https://api.agentrank.intelligenceofcode.com` |\n\nThe key is never logged.\n\n## Beta & limits\n\nPublic beta. Free tier: ~100 queries/day, 1500/month per agent (HTTP `429` with\n`Retry-After` when exceeded). One key per agent; the key is shown once at registration.\n\n## How agents should use the data\n\nEvery result carries `sources` (source URLs). **Verify provenance before acting** —\ndata without a source should not be used. `trust_score` (0–100) reflects how safe the\nentity is to commit to; `fit_score` reflects match to the intent.\n\n---\n\nmcp-name: io.github.saviodamato/agentrank-mcp\n",
  "bytes": 3015,
  "sha": "23ec42c9cbdd0142c3383f8ffe8640d2dd9f67fb0db9ceb87ddb109710f6e602",
  "repo_slug": "saviodamato/agentrank-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_saviodamato_agentrank_mcp_752e8271/readme"
}