winston-ai

Winston AI MCP Server provides a centralized interface to access multiple AI assessment tools through the MCP (Model Context Protocol). It exposes capabilities for AI text detection, AI image detection, plagiarism checking, and text comparison, all accessible via a unified JSON-RPC API. You can invoke tools like ai-text-detection, ai-image-detection, plagiarism-detection, and text-compare by calling the server's methods with the appropriate parameters. The server is designed to be used locally or via remote endpoints and expects your Winston AI API key to authenticate requests. Typical usage includes listing available tools, then calling a specific tool with the required arguments (e.g., text to analyze, image URL, or texts to compare). The server also supports environment configuration to securely supply your API key in different environments (e.g., via cloud adapters or IDE integrations).

Prerequisites:

• Node.js 18+ (recommended)
• npm or pnpm

Installation steps:

1. Install dependencies and build the MCP server:

# Clone the repository (if you haven't already)
git clone https://github.com/gowinston-ai/winston-ai-mcp-server.git
cd winston-ai-mcp-server

# Install dependencies
npm install

# Build TypeScript to JavaScript (if applicable in this project)
npm run build

2. Run the MCP server in a local environment:

# Start the MCP server (uses npx to run the package)
env WINSTONAI_API_KEY=your-api-key npx -y winston-ai-mcp

3. (Optional) Run via Docker:

# Build the image (from the repository Dockerfile, if provided)
docker build -t winston-ai-mcp .

# Run the container with your API key
docker run -e WINSTONAI_API_KEY=your_api_key -p 8080:8080 winston-ai-mcp

Prerequisites summary: ensure Node.js 18+ is installed, obtain a Winston AI API key at the official site, and have network access to reach the Winston AI services. The MCP server can be started with npx or integrated into your preferred environment using the provided configuration examples.

Tips and notes:

• Ensure you supply your Winston AI API key via environment variables (e.g., WINSTONAI_API_KEY) to authenticate requests.
• When integrating with IDEs or desktop helpers (Claude/Cursor), you can reference the MCP server with a command and args (npx -y winston-ai-mcp) and pass the API key through env.
• The API endpoints follow JSON-RPC conventions; use tools/list to discover available tools, then tools/call with the appropriate parameters for each tool (ai-text-detection, ai-image-detection, plagiarism-detection, text-compare).
• If you plan to run in production, consider securing the API key and using a reverse proxy with HTTPS termination.
• For Docker usage, remember to map ports if you intend to expose the MCP server externally, and inject environment variables for authentication.
• Check the Winston AI MCP documentation for the exact parameter shapes for each tool (e.g., required fields like text, url, images, and apiKey).