image

This MCP server provides image analysis capabilities. It accepts input in two forms: an image URL or a local file path. When fed with an image URL, the server retrieves the image and analyzes its content, producing a descriptive analysis and details about objects, scenes, and other notable features. When supplied with a local file path, the server reads the image from the file system and returns a similar analysis. The underlying model used is a GPT-4o-mini variant, enabling high-precision recognition and natural-language descriptions. To use it in a workflow, you configure the MCP client (e.g., Cline or Claude Desktop) to point at the server command and provide the OPENAI_API_KEY as an environment variable. The server exposes two tools: analyze_image for URL-based analysis and analyze_image_from_path for local-file-based analysis. When requesting analysis from a path, ensure the client environment can access and pass a valid filesystem path compatible with the host OS.

Prerequisites:

• Node.js (recommended LTS) and npm installed on your system
• Git installed for cloning the repository
• An OpenAI API key to set OPENAI_API_KEY

1. Clone the repository or install via npm if available:

   git clone https://github.com/champierre/image-mcp-server.git cd image-mcp-server

2. Install dependencies:

   npm install

3. Build the TypeScript project (if applicable):

   npm run build

4. Run the server locally (example):

   OPENAI_API_KEY=your_openai_api_key npm run dev

5. Optional: package for distribution (if you publish to npm):

   npm publish

Note: If you encounter a TypeScript TS7016 error about mime-types during build, it is a type-checking issue that does not affect runtime; install types as a dev dependency if desired:

npm install --save-dev @types/mime-types

Environment variables and configuration tips:

• OPENAI_API_KEY must be set to a valid OpenAI API key for image analysis.
• When integrating with MCP clients (e.g., Cline or Claude Desktop), point the server configuration to the built dist/index.js script and supply OPENAI_API_KEY via env.
• Ensure the path to dist/index.js is correct for your deployment (e.g., /usr/local/bin/image-mcp-server/dist/index.js).
• If the server is behind a firewall or requires proxy settings, configure network access accordingly for image retrieval from URLs.
• For local file analysis, ensure the hosting environment permits file system access to the provided path.
• If you update dependencies or change the build output path, remember to update the MCP config accordingly.