Skip to main content
Query Contextual AI documentation in real-time from your favorite AI tools using the Model Context Protocol (MCP). Ask questions about our APIs, SDKs, Agent Composer, and more—without leaving your development environment.

​
Overview

Our documentation is available as an MCP server, allowing AI applications to search and retrieve information directly. This means you can:
  • Ask questions about Contextual AI APIs and features while coding
  • Get code examples from our documentation in context
  • Search across all docs without switching browser tabs

​
MCP Server URL

https://docs.contextual.ai/mcp

​
Setup Instructions

​
Claude Desktop

  1. Navigate to Claude’s Connectors settings page
  2. Select Add custom connector
  3. Enter the details:
    • Name: Contextual AI Docs
    • URL: https://docs.contextual.ai/mcp
  4. Click Add
  5. During chat, use the attachments button (plus icon) to access the docs

​
Claude Code

Run this command in your terminal: 
claude mcp add --transport http contextual-docs https://docs.contextual.ai/mcp
Verify the installation: 
claude mcp list

​
Cursor

  1. Open the command palette (Cmd+Shift+P on Mac, Ctrl+Shift+P on Windows)
  2. Search for MCP: Open User Configuration
  3. Add the following to your mcp.json:
{
  "mcpServers": {
    "contextual-docs": {
      "url": "https://docs.contextual.ai/mcp"
    }
  }
}
  1. Restart Cursor

​
VS Code

Create a .vscode/mcp.json file in your project: 
{
  "servers": {
    "contextual-docs": {
      "type": "http",
      "url": "https://docs.contextual.ai/mcp"
    }
  }
}

​
Usage Examples

Once connected, you can ask questions like:
  • “How do I create a datastore using the Python SDK?”
  • “What are the parameters for the rerank API?”
  • “Show me an example of Agent Composer YAML configuration”
  • “How do I configure retrieval settings for my agent?”
Your AI tool will search the Contextual AI documentation and provide answers with relevant context.

​
Notes

  • The MCP server provides access to all public documentation pages
  • Multiple MCP servers can be connected simultaneously—context is only consumed when actively searched
  • The AI tool automatically determines when to search based on your query