Mcp Server Python

What is Mcp Server Python about?

Mcp Server Python offers a Model Context Protocol (MCP) server that connects to Inkeep's API to perform retrieval‑augmented generation (RAG) on product documentation and other knowledge assets. It allows client applications, such as Claude Desktop, to query Inkeep content in a conversational manner.

How to use Mcp Server Python?

1. Clone the repository and set up a virtual environment using uv.
2. Install dependencies with uv pip install -r pyproject.toml.
3. Obtain an Inkeep API key from the Inkeep Dashboard → Projects → Integrations.
4. Configure your MCP client (e.g., Claude Desktop) by adding an entry to claude_desktop_config.json that points to the server's directory, command, and required environment variables.
5. Run the server via the configured command; the client will forward queries to the server, which in turn calls the Inkeep API and returns relevant results.

Key Features of Mcp Server Python

• Seamless integration with Inkeep's RAG endpoints.
• Configurable via environment variables for API base URL, key, model, tool name, and description.
• Works with the Model Context Protocol, allowing language model clients to treat the server as a built‑in tool.
• Simple Python setup using uv for virtual environments and dependency management.
• Example configuration provided for Claude Desktop.

Use Cases of Mcp Server Python

• Building AI assistants that can answer product‑specific questions using up‑to‑date documentation.
• Enhancing customer support bots with context‑aware retrieval of knowledge base articles.
• Enabling internal tools that need to surface technical content from Inkeep without writing custom API calls.
• Integrating RAG capabilities into existing AI workflows that support MCP.

FAQ from the Mcp Server Python

Q: Do I need an Inkeep account? A: Yes, an active Inkeep account and a project with API access are required.

Q: Which command should I use to run the server? A: The recommended command is uv run -m inkeep_mcp_server within the project directory, as shown in the provided MCP client configuration.

Q: Can I use a different language model? A: The server defaults to the inkeep-rag model, but you can change INKEEP_MCP_TOOL_NAME and other environment variables to target a different model if supported by Inkeep.

Q: How do I set the path to uv? A: On macOS/Linux run which uv; on Windows run where uv, then use the full path in the command field if necessary.

Q: Is there a Docker image available? A: The repository does not currently provide a Docker setup; you can containerize it manually by copying the repository and installing dependencies inside a container.

mcp-server-python

Inkeep MCP Server powered by your docs and product content.

Dependencies

• An account on Inkeep to manage and provide the RAG
• uv Python project manager

Local Setup

git clone https://github.com/inkeep/mcp-server-python.git
cd mcp-server-python
uv venv
uv pip install -r pyproject.toml

Note the full path of the project, referred to as <YOUR_INKEEP_MCP_SERVER_ABSOLUTE_PATH> in a later step.

Get an API key

1. Log in to the Inkeep Dashboard
2. Navigate to the Projects section and select your project
3. Open the Integrations tab
4. Click Create Integration and choose API from the options
5. Enter a Name for your new API integration.
6. Click on Create
7. A generated API key will appear that you can use to authenticate API requests.

We'll refer to this API key as the <YOUR_INKEEP_API_KEY> in later steps.

Add to your MCP client

Follow the steps in this guide to setup Claude Dekstop.

In your claude_desktop_config.json file, add the following entry to mcpServers.

{
    "mcpServers": {
        "inkeep-mcp-server": {
            "command": "uv",
            "args": [
                "--directory",
                "<YOUR_INKEEP_MCP_SERVER_ABSOLUTE_PATH>",
                "run",
                "-m",
                "inkeep_mcp_server"
            ],
            "env": {
                "INKEEP_API_BASE_URL": "https://api.inkeep.com/v1",
                "INKEEP_API_KEY": "<YOUR_INKEEP_API_KEY>",
                "INKEEP_API_MODEL": "inkeep-rag",
                "INKEEP_MCP_TOOL_NAME": "search-product-content",
                "INKEEP_MCP_TOOL_DESCRIPTION": "Retrieves product documentation about Inkeep. The query should be framed as a conversational question about Inkeep."
            }
        },
    }
}

You may need to put the full path to the uv executable in the command field. You can get this by running which uv on MacOS/Linux or where uv on Windows.