devmentorai-server

AI-powered DevOps mentoring and writing assistant backend, powered by GitHub Copilot.

Part of the DevMentorAI project.

Quick Start

npx devmentorai-server

That's it! The server starts in the background on port 3847.

Installation

Run without installing (recommended)

npx devmentorai-server

Install globally

npm install -g devmentorai-server
devmentorai-server

Long-Running Usage (No Interruptions)

npx devmentorai-server is supported and starts the backend in background mode, but for long-running sessions (hours/days) you should use a process supervisor.

Option 1: Global install + PM2 (recommended)

npm install -g devmentorai-server pm2
pm2 start "devmentorai-server start --foreground" --name devmentorai-server
pm2 save
pm2 status

Option 2: systemd (Linux)

Create a user service that runs:

devmentorai-server start --foreground

Then enable restart policies (Restart=always) so the backend auto-recovers if it crashes.

Commands

| Command | Description | |---------|-------------| | devmentorai-server start | Start the server in background (default) | | devmentorai-server stop | Stop the running server | | devmentorai-server status | Show server status (PID, port, health) | | devmentorai-server logs | View recent server logs | | devmentorai-server doctor | Check system requirements |

Options

| Option | Description | |--------|-------------| | --port <port> | Port to listen on (default: 3847) | | --foreground, -f | Run in foreground (don't daemonize) | | --lines, -n | Number of log lines to show (default: 50) | | --help, -h | Show help message | | --version, -v | Show version |

Examples

# Start server on a custom port
devmentorai-server start --port 4000

# Run in foreground (useful for debugging)
devmentorai-server start --foreground

# Check if everything is set up correctly
devmentorai-server doctor

# View last 100 lines of logs
devmentorai-server logs --lines 100

Requirements

• Node.js >= 20.0.0
• GitHub Copilot CLI (optional — server runs in mock mode without it)

How It Works

The server runs as a background process and stores its data in ~/.devmentorai/:

~/.devmentorai/
├── devmentorai.db      # SQLite database (sessions, messages)
├── server.pid          # PID of running server
├── config.json         # User configuration
├── logs/
│   └── server.log      # Server logs
└── images/             # Session image thumbnails

API Endpoints

| Endpoint | Description | |----------|-------------| | GET /api/health | Health check | | GET /api/sessions | List sessions | | POST /api/sessions | Create a new session | | POST /api/chat/:sessionId | Send a message | | GET /api/models | List available models |

Troubleshooting

Server won't start

1. Run devmentorai-server doctor to check requirements
2. Check logs: devmentorai-server logs
3. Try foreground mode: devmentorai-server start -f

Server stops unexpectedly

1. Check the latest backend logs: devmentorai-server logs --lines 200
2. Re-run in foreground to capture runtime errors directly: devmentorai-server start -f
3. For continuous uptime, run under PM2 or systemd (see Long-Running Usage)

Port already in use

devmentorai-server start --port 4000

Copilot not connected

The server requires GitHub Copilot CLI to be installed and authenticated. Without it, the server runs in mock mode with simulated responses.

# Install Copilot CLI (if not installed)
npm install -g @github/copilot-cli

# Authenticate
github-copilot auth

Reset data

rm -rf ~/.devmentorai
devmentorai-server start

Development

This package is part of the DevMentorAI monorepo.

git clone https://github.com/BOTOOM/devmentorai.git
cd devmentorai
pnpm install
pnpm dev:backend

Support

Found a bug or have a feature request? Open an issue on GitHub.

License

MIT © BOTOOM