JiwuChat

JiwuChat is a cross-platform chat application built with a Nuxt/Tauri frontend and a backend core that can be self-hosted. The MCP server for JiwuChat in this repository is provided as a Docker image (jiwuchat-core) which encapsulates the backend services such as user management, real-time messaging, AI integrations, and cross-device synchronization. To use the MCP server, you deploy the backend image and expose its API endpoints so the frontend or other services can communicate with it. Once the backend is running, you can point the JiwuChat frontend to the backend API to enable login, messaging, group chats, voice/video calls, and AI features like AI chat bots and AI shopping integrations. The server is designed to work with the existing self-hosted backend (jiwu-chat-core) and can be deployed with a single docker run command. Ensure your environment can reach any required AI model services and that network ports align with your frontend configuration.

Prerequisites:

• Docker installed and running on your host
• Basic familiarity with Docker image deployment

Step-by-step:

1. Pull and run the JiwuChat core image as the MCP server: docker run -i jiwuchat-core

2. If you need to bind to specific ports or pass environment variables (e.g., API keys, database settings), customize the docker run command: docker run -i -p 8000:8000 -e AI_TOKEN=your_token jiwuchat-core

3. Verify the container is running: docker ps docker logs <container_id>

4. Configure the frontend (JiwuChat UI) or any consuming services to point to the backend API endpoint exposed by the container (e.g., http://localhost:8000/api). If you are running in a production environment, replace localhost with your server address and ensure TLS/HTTPS is configured appropriately.

5. Optional: For persistent data, mount volumes to preserve databases/files: docker run -i -p 8000:8000 -v jiwuchat_data:/data jiwuchat-core

Tips:

• If you already have jiwu-chat-core dependencies or services (databases, AI models), ensure the container has network access to those endpoints.
• Common issues include network connectivity between frontend and backend, misconfigured API base URLs, and missing environment variables for AI model providers. Check container logs for hints.
• If you need to restart the MCP server, use docker restart <container_id>.
• For production deployments, consider using docker-compose or Kubernetes to manage multiple services (frontend, backend, database, AI services) and to handle scaling and health checks.
• If you require specific environment variables (e.g., DI services, AI model keys), document them in the env section of your deployment YAML or docker run command.