azure-ai-vision-face

This MCP server implements a Face Liveness Detection workflow to prove presence during AI agent workflows. It exposes a Node.js based high-level tool that can be integrated into Claude or other MCP clients via a standard mcpServers configuration. The server relies on Microsoft Azure Face APIs for liveness verification and expects environment variables for endpoint, key, and related settings. To use it, configure the MCP entry for the server (as shown in the example) and ensure the provided SESSION_IMAGE_DIR exists so session images can be saved. When deployed, you can invoke the liveness verification as part of a broader automation flow, then use the returned results to decide the next steps in your agent pipeline.

Prerequisites: Node.js and npm installed on your system. If you are using the TypeScript example from the repository, follow these steps:

1. Navigate to the liveness-server-typescript folder:

   • cd liveness-server-typescript

2. Install dependencies:

   • npm install --save-dev typescript @types/node
   • npm install

3. Build the TypeScript project:

   • npm run build

4. In the build output, you should have an index.js (or stdio.js) that serves as the entrypoint for the MCP integration. Use this path in your MCP configuration (e.g., YOUR_PATH/build/stdio.js).

5. Ensure your environment variables are set in your MCP config as shown in the mcp_config example.

Tips and common issues:

• Ensure SESSION_IMAGE_DIR is set; otherwise session images will not be saved.
• Verify_IMAGE_FILE_NAME can be used to switch to verify mode by pointing to a stored verify image.
• The example uses a Node.js entrypoint (stdio.js) produced by building the TypeScript source; adjust the path if your setup differs.
• Replace placeholder values in FACEAPI_ENDPOINT, FACEAPI_KEY, and FACEAPI_WEBSITE with your actual Azure Face service credentials.
• When testing locally, ensure the path in args points to the built entrypoint file that launches the MCP server.

Configuration options:

• command: node
• args: ["YOUR_PATH/build/stdio.js"]
• env: FACEAPI_ENDPOINT, FACEAPI_KEY, FACEAPI_WEBSITE, SESSION_IMAGE_DIR