Skip to content

SELF_HOSTING.md

Latest commit

 

History

History
227 lines (162 loc) · 5.39 KB

SELF_HOSTING.md

File metadata and controls

227 lines (162 loc) · 5.39 KB

Self-Hosting Coderrr Backend

Most users don't need this! The default hosted backend works great. But if you want to run your own backend for privacy, customization, or development, follow this guide.

Why Self-Host?

  • Privacy: Keep your code and AI interactions private
  • Customization: Use different AI models or providers
  • Development: Contribute to the backend or test changes
  • Cost Control: Use your own API keys and control costs

Prerequisites

  • Python 3.8 or higher
  • pip (Python package manager)
  • An AI provider API key (GitHub Models, Mistral AI, etc.)

Installation Steps

1. Clone the Repository

git clone https://github.com/Akash-nath29/Coderrr.git
cd Coderrr

2. Set Up Python Virtual Environment

Windows:

python -m venv env
.\env\Scripts\Activate.ps1

Linux/Mac:

python3 -m venv env
source env/bin/activate

3. Install Python Dependencies

pip install -r backend/requirements.txt

4. Configure Backend

Create backend/.env file:

cp backend/.env.example backend/.env

Edit backend/.env with your API credentials:

# Choose ONE authentication method:

# Option 1: GitHub Models (Free tier available)
GITHUB_TOKEN=ghp_your_github_token_here

# Option 2: Mistral AI Direct
# MISTRAL_API_KEY=your_mistral_api_key_here

# Model Configuration
MISTRAL_ENDPOINT=https://models.inference.ai.azure.com
MISTRAL_MODEL=mistral-large-2411

# Server Configuration
TIMEOUT_MS=120000

Getting API Keys

GitHub Models (Recommended for Free Tier):

  1. Go to https://github.com/settings/tokens
  2. Click "Generate new token" → "Generate new token (classic)"
  3. Select scopes: read:user (minimal)
  4. Copy the token and paste in GITHUB_TOKEN

Mistral AI:

  1. Go to https://console.mistral.ai/
  2. Sign up and navigate to API Keys
  3. Create a new API key
  4. Copy and paste in MISTRAL_API_KEY

5. Start the Backend

Development Mode (with auto-reload):

cd backend
uvicorn main:app --reload --port 5000

Production Mode:

cd backend
uvicorn main:app --host 0.0.0.0 --port 5000 --workers 4

The backend will be available at http://localhost:8000

6. Configure CLI to Use Local Backend

Create ~/.coderrr/.env:

Windows:

mkdir $HOME\.coderrr
echo CODERRR_BACKEND=http://localhost:8000 > $HOME\.coderrr\.env

Linux/Mac:

mkdir -p ~/.coderrr
echo "CODERRR_BACKEND=http://localhost:8000" > ~/.coderrr/.env

7. Test It

coderrr exec "Create a hello world script"

Production Deployment

For production deployment, see our Deployment Guide which covers:

  • Docker deployment
  • Cloud hosting (AWS, Azure, GCP)
  • Vercel/Netlify deployment
  • PM2 process management
  • Nginx reverse proxy setup

Troubleshooting

Backend won't start

Error: ModuleNotFoundError: No module named 'fastapi'

  • Solution: Make sure you activated the virtual environment and ran pip install -r backend/requirements.txt

Error: Address already in use

  • Solution: Port 5000 is already taken. Either kill the process using port 5000 or use a different port: 
    uvicorn main:app --port 5001
    Then update CODERRR_BACKEND=http://localhost:5001

CLI can't connect to backend

Error: Failed to communicate with backend: ECONNREFUSED

  • Check backend is running: curl http://localhost:8000 should return a JSON response
  • Check ~/.coderrr/.env has correct CODERRR_BACKEND URL
  • Check firewall isn't blocking port 5000

API authentication errors

Error: 401 Unauthorized

  • Verify your API key is correct in backend/.env
  • For GitHub token: check it has required permissions
  • For Mistral: verify key is active on their console

Advanced Configuration

Using Different AI Models

Edit backend/.env:

# Use GPT-4 via Azure
MISTRAL_ENDPOINT=https://your-azure-endpoint.openai.azure.com
MISTRAL_MODEL=gpt-4

# Use Claude via custom endpoint
MISTRAL_ENDPOINT=https://api.anthropic.com
MISTRAL_MODEL=claude-3-opus

Note: You may need to modify backend/main.py for providers other than Mistral/GitHub Models.

Increasing Timeout

For large requests:

TIMEOUT_MS=300000  # 5 minutes

Rate Limiting

Add to backend/main.py:

from fastapi import FastAPI
from slowapi import Limiter, _rate_limit_exceeded_handler
from slowapi.util import get_remote_address
from slowapi.errors import RateLimitExceeded

limiter = Limiter(key_func=get_remote_address)
app = FastAPI()
app.state.limiter = limiter
app.add_exception_handler(RateLimitExceeded, _rate_limit_exceeded_handler)

@app.post("/chat")
@limiter.limit("10/minute")
async def chat(request: Request, body: ChatRequest):
    # ... existing code

Security Considerations

  • Never commit .env files to git
  • Use strong API keys and rotate them regularly
  • Run backend on localhost only unless you need remote access
  • Enable firewall if exposing backend to internet
  • Use HTTPS in production with reverse proxy
  • Monitor API usage to prevent unexpected costs

Need Help?

Remember: Most users don't need to self-host! The default hosted backend at https://coderrr-backend.vercel.app works great for normal use.