Claude Context

Sign up at cloud.zilliz.com and create a free Serverless cluster. Copy the Public Endpoint (your MILVUS_ADDRESS) and generate an API key (your MILVUS_TOKEN) from the cluster detail page.

Log in to platform.openai.com, navigate to API Keys, and create a new secret key. This key is used to generate vector embeddings for semantic code search.

Run the following command to register claude-context as an MCP server with your credentials. Replace the placeholder values with your actual keys.

claude mcp add claude-context \
  -e OPENAI_API_KEY=sk-your-openai-api-key \
  -e MILVUS_ADDRESS=your-zilliz-cloud-public-endpoint \
  -e MILVUS_TOKEN=your-zilliz-cloud-api-key \
  -- npx @zilliz/claude-context-mcp@latest

For Claude Desktop, Cursor, Windsurf, or VS Code, add the server block to your MCP config file (e.g., claude_desktop_config.json or .cursor/mcp.json).

{
  "mcpServers": {
    "claude-context": {
      "command": "npx",
      "args": ["-y", "@zilliz/claude-context-mcp@latest"],
      "env": {
        "OPENAI_API_KEY": "sk-your-openai-api-key",
        "MILVUS_ADDRESS": "your-zilliz-cloud-public-endpoint",
        "MILVUS_TOKEN": "your-zilliz-cloud-api-key"
      }
    }
  }
}

Open your AI assistant and ask it to index the codebase. The server will crawl the project directory and build a hybrid search index in Zilliz Cloud. Large repos may take a few minutes.

Once indexing is complete, ask natural-language questions about your codebase. The server exposes search_code to retrieve relevant snippets, get_indexing_status to monitor progress, and clear_index to remove an existing index.