home / mcp / dock ai mcp server
Provides MCP endpoints to interact with real-world entities like restaurants and hotels, enabling discovery, reservations, and availability actions.
Configuration
View docs{
"mcpServers": {
"dock-ai-mcp": {
"url": "https://mcp.dockai.co/mcp"
}
}
}Dock AI MCP lets your agents discover which MCP endpoints can interact with a real-world entity, such as a restaurant or hotel, by querying a registry. It simplifies connecting to the right MCP servers for specific entities and actions like reservations or availability checks.
You use an MCP client to query the Dock AI MCP registry and obtain a list of MCP endpoints that support the entity you care about. For example, you can resolve the domain of a restaurant to learn which MCPs offer reservations or availability data, then connect your agent to the appropriate MCP server and perform the desired action.
Prerequisites you need before running an MCP server locally:
Option A: Run the hosted MCP service (no installation required) then point your client to the MCP URL.
Option B: Self-host locally using one of the supported runtimes.
# Use the runtime via uvx
uvx dock-ai-mcp
# Or install the Python package and run the server
pip install dock-ai-mcp
dock-ai-mcpHosted version: you can use the MCP service at the following address without installing anything on your side: https://mcp.dockai.co/mcp. This provides immediate access to Dock AI MCP endpoints for real-world entities.
Self-hosting notes: when you run locally, the server starts and listens on http://0.0.0.0:8080/mcp by default. You can customize this if your deployment requires a different host or port.
Resolve a domain to its MCP endpoints, returning the entity details and the MCP providers with their capabilities.