For AI Agents
Use the Dock AI MCP server to discover MCP endpoints for real-world entities.
Quick Start
Add Dock AI to your Claude Desktop configuration:
{
"mcpServers": {
"dock-ai": {
"url": "https://mcp.dockai.co/mcp"
}
}
}Config file location: ~/Library/Application Support/Claude/claude_desktop_config.json
Remote URL: https://mcp.dockai.co/mcp
No installation required. Just add the URL to your MCP client.
How it works
- User asks to interact with a business (e.g., "Book a table at example-restaurant.com")
- AI calls
resolve_domain("example-restaurant.com") - Dock AI returns the MCP endpoint(s) that can handle this entity
- AI suggests installing the provider's MCP to complete the action
Tools
resolve_domain
Resolve a domain to its MCP endpoints.
| Parameter | Type | Description |
|---|---|---|
| domain | string | Domain to resolve (e.g., "example-restaurant.com") |
Response
{
"domain": "example-restaurant.com",
"entities": [
{
"name": "Acme Bistro",
"path": "/",
"category": "restaurant",
"verification_level": 2,
"mcps": [
{
"provider": "booking-provider",
"endpoint": "https://mcp.booking-provider.com",
"capabilities": ["reservations", "availability"]
}
]
}
],
"claude_desktop_config": {
"mcpServers": {
"booking-provider": { "url": "https://mcp.booking-provider.com/mcp" }
}
}
}The response includes a claude_desktop_config snippet that the AI can suggest to the user for installing the provider's MCP.
Self-hosting
You can run your own instance of the Dock AI MCP server:
# Run with uvx
uvx dock-ai-mcp
# Or deploy to Vercel
# See: https://github.com/dock-ai/mcp