Sudosu 🚀

Your AI Coworker - Right in Your Terminal

Sudosu gives you AI coworkers that can actually get work done. Unlike chatbots that just talk, these teammates can read your files, write code, create documents, connect to all your tools (Gmail, Calendar, GitHub, Linear, Slack), and run commands - all while you stay in control.

No more hopping between tools. Your AI coworker does it all.

Installation

pip install sudosu

Quick Start

Zero configuration required. Just install and run:

# Install
pip install sudosu

# Start using immediately - no setup needed!
sudosu

That's it! Sudosu works out of the box with our hosted backend.

Example Tasks

# Start interactive session
sudosu

# Then just ask:
> Summarize the unread emails in my inbox
> Create a Linear ticket for the bug we discussed
> Check my calendar for tomorrow and draft a prep email
> Go through #product-team slack and summarize yesterday's messages

Connect Your Tools

# Inside sudosu, connect integrations:
/connect gmail      # Connect Gmail
/connect slack      # Connect Slack  
/connect linear     # Connect Linear
/connect github     # Connect GitHub
/connect notion     # Connect Notion

Self-Hosting (Optional)

Want to run your own backend? You can self-host the Sudosu backend with Docker:

Quick Self-Host

# Pull the Docker image
docker pull akash07/sudosu-cli-image:latest

# Run with your own configuration
docker run -d -p 8000:8000 \
  -e GOOGLE_API_KEY=your-google-api-key \
  -e COMPOSIO_API_KEY=your-composio-api-key \
  -e DATABASE_URL=postgresql://user:pass@host:5432/db \
  -e REDIS_URL=redis://host:6379 \
  -e RESEND_API_KEY=your-resend-api-key \
  akash07/sudosu-cli-image:latest

Full Stack with Docker Compose

For a complete setup with Redis and background worker, create a docker-compose.yml file:

use this compose code : https://gist.github.com/csakash/d26619add372013af6d0e7b119af28c0

Create a .env file with your credentials:

# Create .env file
cat > .env << EOF
GOOGLE_API_KEY=your-google-api-key-here
COMPOSIO_API_KEY=your-composio-api-key-here
DATABASE_URL=postgresql://user:password@host:5432/database
RESEND_API_KEY=your-resend-api-key-here
LOG_LEVEL=INFO
EOF

# Start all services
docker-compose up -d

Connect CLI to Your Self-Hosted Backend

After starting your backend, configure the CLI:

# Set development mode to use local backend
export SUDOSU_MODE=dev
export SUDOSU_BACKEND_URL=http://localhost:8000

# Or configure within CLI
sudosu
> /config mode dev
> /config backend http://localhost:8000

Requirements for Self-Hosting

• PostgreSQL Database: Any provider (Neon, Supabase, local, etc.)
• Redis: Included in docker-compose or use external instance
• API Keys:
  • Google API Key (Gemini): Get here
  • Composio API Key: Get here
  • Resend API Key: Get here

For detailed self-hosting instructions, see the backend repository.

Features

• 🚀 Zero Config: Install and run - works immediately with hosted backend
• 🤖 AI Coworkers: Create specialized coworkers with specific personalities and capabilities
• 🔌 Tool Integrations: Connect to Gmail, Calendar, GitHub, Linear, Slack, Notion, and more
• 📝 File Operations: Coworkers can read and write files in your repository
• 🔄 Real-time Streaming: See responses as they're generated
• 🔒 Local Execution: File operations happen on your machine, keeping data secure
• ⚡ Action-Oriented: Your coworkers don't just answer questions — they take action

Commands

Configuration (Optional)

Sudosu works out of the box, but you can customize it:

Environment Modes

# Development mode (local backend)
export SUDOSU_MODE=dev
sudosu

# Production mode (default - uses hosted backend)
export SUDOSU_MODE=prod
sudosu

# Or switch within CLI
/config mode dev   # Switch to development
/config mode prod  # Switch to production

Configuration Files

Sudosu stores minimal global config in ~/.sudosu/:

~/.sudosu/
└── config.yaml     # API keys, mode settings, user ID

Project-level configuration is created automatically when you run sudosu in any folder:

your-project/
└── .sudosu/
    ├── AGENT.md    # Your customizable AI assistant prompt
    ├── agents/     # Custom agents created with /agent create
    └── context.md  # (optional) Project context for all agents

Edit .sudosu/AGENT.md to customize how your AI assistant behaves in that project.

Requirements

• Python 3.10+
• Internet connection (for hosted backend)

Links

• Website: trysudosu.com
• Issues: GitHub Issues

License

MIT