cortex-cloud-docs

This MCP server provides fast search access to Cortex Cloud documentation and API references for Claude and other MCP-compatible clients. After deployment, you can index the documentation and API references locally and then query them via JSON-RPC. The server exposes tools to index Cortex Cloud docs, index the API docs, search within the docs, and retrieve index statistics. The available tools are designed to be called from your MCP client, such as Claude Desktop, via the standard MCP methods and a simple set of function signatures described in the README.

Prerequisites:

• Python 3.12 or higher
• uv package manager (uv) or uvx for PyPI integration via Claude Desktop

Option A: Install via Smithery (recommended for Claude Desktop users)

1. Install the MCP server for Claude Desktop:

   npx -y @smithery/cli install @clarkemn/cortex-cloud-docs-mcp-server --client claude
   

2. In Claude Desktop, add the server configuration pointing to the PyPI package via uvx as shown in the usage example.

Option B: Development installation (clone and run locally) Prerequisites:

• Python 3.12+ installed
• uv (Astral uv) installed

Install uv (if not already):

curl -LsSf https://astral.sh/uv/install.sh | sh

Clone and set up:

git clone https://github.com/clarkemn/cortex-cloud-docs-mcp-server.git
cd cortex-cloud-docs-mcp-server
uv sync

Run the server locally (development):

uv run python server.py

Note: The repository uses a Python-based MCP server. If you plan to run directly, ensure you are in the project root and that dependencies are installed via uv sync. You can also install from PyPI and run via uvx for Claude Desktop integration as described above.

Tips and common issues:

• If the server does not start in Claude Desktop, ensure uv is installed and available in PATH, and verify the path to the project directory if using local development.
• Before performing searches, run index_cortex_docs(max_pages) or index_cortex_api_docs(max_pages) to populate the index; otherwise, search results may be empty.
• Use search_all_docs(query) to search across all indexed documentation when you are unsure whether a term exists in the docs or API references.
• If dependencies are missing after cloning, run uv sync to install them.
• The environment variables (env) can be used to configure API keys or feature flags if needed; start with an empty env for standard operation and add placeholders like {"ENV_VAR": "description"} as needed for your deployment.
• The available tools include: index_cortex_docs, index_cortex_api_docs, search_cortex_docs, search_cortex_api_docs, search_all_docs, get_index_status.