- Step1: Clone the repository and navigate into the project directory.
- Step2: Install dependencies with npm install or yarn install.
- Step3: Add your Semantic Scholar API key and OpenAI (or other LLM) credentials to .env.
- Step4: Configure MCP chains in config/mcp.config.js to define fetch, summarize, and QA steps.
- Step5: Start the server with npm start or node server.js.
- Step6: Send HTTP requests to the provided endpoints (e.g., /api/v1/query) with your research queries.
- Step7: Parse and integrate the JSON responses into your application workflows.