@scoutscore/mcp-server

MCP (Model Context Protocol) server for ScoutScore - Agent Trust Intelligence.

Gives Claude Desktop, Cursor, ChatGPT, and any MCP-compatible client the ability to look up trust scores, fidelity checks, and ecosystem data for AI agent services.

Tools

| Tool | Description | |------|-------------| | check_trust | Look up trust score, dimensions, flags, and recommendation for a service domain | | check_fidelity | Check whether a service's actual responses match its advertised contract | | get_ecosystem_stats | Get aggregate stats about the ScoutScore ecosystem (total services, pricing, schema coverage) | | search_services | Search and browse the service leaderboard with keyword and category filters | | scan_skill | Scan a GitHub-hosted AI skill/plugin/tool for trust and safety |

Installation

Claude Desktop

Add to your claude_desktop_config.json:

macOS: ~/Library/Application Support/Claude/claude_desktop_config.json Windows: %APPDATA%\Claude\claude_desktop_config.json Linux: ~/.config/Claude/claude_desktop_config.json

{
  "mcpServers": {
    "scoutscore": {
      "command": "npx",
      "args": ["-y", "@scoutscore/mcp-server"]
    }
  }
}

Cursor

Add to your .cursor/mcp.json (project-level) or ~/.cursor/mcp.json (global):

{
  "mcpServers": {
    "scoutscore": {
      "command": "npx",
      "args": ["-y", "@scoutscore/mcp-server"]
    }
  }
}

ChatGPT

ChatGPT supports MCP servers via the desktop app. Go to Settings > MCP Servers > Add Server and enter:

• Name: ScoutScore
• Command: npx
• Arguments: -y @scoutscore/mcp-server

Claude Code

Add to your project's .mcp.json:

{
  "mcpServers": {
    "scoutscore": {
      "command": "npx",
      "args": ["-y", "@scoutscore/mcp-server"]
    }
  }
}

Manual / Other Clients

Run the server directly:

npx @scoutscore/mcp-server

Or install globally:

npm install -g @scoutscore/mcp-server
scoutscore-mcp

Example Prompts

Once configured, you can ask your AI assistant things like:

• "Is api.example.com trustworthy? Check its ScoutScore."
• "What's the fidelity score for myservice.xyz?"
• "How many AI agent services are in the ScoutScore ecosystem?"
• "Search for image-related AI services on ScoutScore."
• "Scan the GitHub repo owner/mcp-server for trust and safety."
• "Compare the top 5 AI & ML services on the ScoutScore leaderboard."

Development

# Install dependencies
cd packages/mcp-server
npm install

# Build
npm run build

# Run tests
npm test

# Run live API integration tests
npm run test:live

# Start the server (for testing with an MCP client)
node dist/index.js

How It Works

This MCP server wraps the public ScoutScore API at https://scoutscore.ai. No API key is required - all endpoints are free tier. The server communicates over stdio using the Model Context Protocol, making it compatible with any MCP client.

License

MIT