Offern an Open AI compatible back-end to and host a multi agentic graph.

This project implements an Open AI API that:

• Communicates with a custom React agent
• Utilizes an MCP server for dynamic tool loading and execution
• Built with TypeScript for type safety and better development experience

• Node.js (Latest LTS version recommended)
• npm/yarn
• OpenAI API key
• MCP server access

Create a .env file in the root directory with the following variables:

PORT=3000
OPENAI_PORT=3004
OPENAI_API_KEY=your_openai_api_key

npm install

• npm start: Run the production build from the dist directory
• npm run build: Compile TypeScript to JavaScript
• npm run dev: Run development server with hot-reload
• npm run debug: Run with debugging enabled
• npm run lint: Run ESLint checks
• npm run lint:fix: Fix ESLint issues automatically

• npm run package: Build container image using Podman
• npm run package:run: Run the containerized application with environment variables
• npm run package:tag: Tag the container image for GitHub Container Registry
• npm run push:login-stefan: Login to GitHub Container Registry (requires 1Password CLI)

• @langchain/core, @langchain/community, @langchain/openai: LangChain integration
• @langchain/mcp-adapters: MCP server integration
• dotenv: Environment variable management
• typescript: TypeScript support

• ts-node-dev: TypeScript development server
• eslint: Code linting
• @types/node: TypeScript Node.js types

Run the development server:

npm run dev

The server will restart automatically when you make changes to the source code.

Build and run for production:

npm run build
npm start

1. Build the container:

npm run package

2. Run the container:

npm run package:run

3. Tag and push to GitHub Container Registry:

npm run package:tag
# Login to GitHub Container Registry first
npm run push:login-stefan