> ## Documentation Index
> Fetch the complete documentation index at: https://docs.predicate.io/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Assistance

> Connect Predicate documentation to AI coding assistants via MCP

## MCP Server Integration

Connect AI assistants directly to Predicate documentation for real-time search and reference.

<CodeGroup>
  ```json Cursor theme={null}
  // Open MCP Settings → New MCP Server
  {
    "mcpServers": {
      "predicate": {
        "url": "https://docs.predicate.io/mcp"
      }
    }
  }
  ```

  ```bash Claude Code theme={null}
  claude mcp add --transport http predicate https://docs.predicate.io/mcp
  ```

  ```json Claude Desktop theme={null}
  // Add to ~/.claude/mcp.json
  {
    "mcpServers": {
      "predicate-docs": {
        "transport": "sse",
        "url": "https://docs.predicate.io/mcp",
        "readOnly": true
      }
    }
  }
  ```

  ```json Windsurf theme={null}
  // Settings → MCP Servers → Add Server
  {
    "name": "predicate-docs",
    "url": "https://docs.predicate.io/mcp", 
    "transport": "sse",
    "readOnly": true
  }
  ```

  ```json VS Code theme={null}
  // Add to .continue/config.json
  {
    "experimental": {
      "modelContextProtocol": {
        "servers": [{
          "name": "predicate-docs",
          "url": "https://docs.predicate.io/mcp",
          "transport": "sse"
        }]
      }
    }
  }
  ```
</CodeGroup>

<Tip>
  If you can't use MCP servers copy & paste [llms-full.txt](https://docs.predicate.io/llms-full.txt) directly into your AI conversations
</Tip>