Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.augmentcode.com/llms.txt

Use this file to discover all available pages before exploring further.

Experimental API - Context Connectors is experimental and subject to breaking changes.
For local MCP integration with Claude Desktop, Claude Code, Cursor, or other agents, see Local MCP Server.

​
Prerequisites

​
Steps

​
1. Install

npm install @augmentcode/context-connectors @modelcontextprotocol/sdk

​
2. Set credentials

export AUGMENT_SESSION_AUTH='<session-json>'  # from auggie token print

​
3. Start the server

npx ctxc mcp http -i my-project --port 8080 --api-key "secret"
You should see: 
MCP HTTP server running at http://localhost:8080/mcp

​
4. Test with curl

First, initialize a session: 
curl -X POST http://localhost:8080/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -H "Authorization: Bearer secret" \
  -d '{
    "jsonrpc": "2.0",
    "id": 1,
    "method": "initialize",
    "params": {
      "protocolVersion": "2024-11-05",
      "capabilities": {},
      "clientInfo": { "name": "curl-test", "version": "1.0.0" }
    }
  }' -i
Note the mcp-session-id header in the response. Use it for subsequent requests: 
curl -X POST http://localhost:8080/mcp \
  -H "Content-Type: application/json" \
  -H "Accept: application/json, text/event-stream" \
  -H "Authorization: Bearer secret" \
  -H "mcp-session-id: <session-id-from-initialize>" \
  -d '{
    "jsonrpc": "2.0",
    "id": 2,
    "method": "tools/call",
    "params": {
      "name": "search",
      "arguments": { "query": "authentication" }
    }
  }'
You should see JSON search results.

​
Done!

Your index is accessible via MCP over HTTP at http://localhost:8080/mcp. Remote clients can now search your codebase using the Model Context Protocol over HTTP.

​
Also Works With

Instead of…Try…
localhost only--host 0.0.0.0 to accept external connections
No CORS--cors "*" or --cors "https://myapp.com" for browser clients
Full access--search-only to disable file operations
Local storage--store s3 with CC_S3_BUCKET for shared indexes