StayGreen

Keep your GitHub contribution graph consistently green with AI-powered automated commits! StayGreen automatically tracks your actual coding work every 1 hour and creates meaningful commits.

Features

• 🟩 Automatic commit tracking every 1 hour
• 🤖 AI-powered commit messages using local Ollama models or GPT-4
• 📝 Follows conventional commit format
• 📊 Daily organized summaries
• 🔒 Secure GitHub OAuth authentication
• 🎯 Zero configuration needed
• 🔋 Local AI support with Ollama

Why StayGreen?

• Show Real Progress: Automatically captures your actual coding work
• Smart Commits: Uses AI (local or cloud) to generate meaningful commit messages
• Stay Consistent: Never miss showing your contributions
• Zero Effort: Works automatically after login
• Privacy First: Option to use local AI models

Requirements

• GitHub Account
• Git installed and configured
• VS Code version 1.50.0 or higher
• Either:
  • Ollama installed (for local AI processing)
  • OpenAI API Key (for cloud-based processing)

Installing Ollama

macOS

# Using Homebrew
brew install ollama

# Start Ollama
ollama serve

Linux

# Install using curl
curl -fsSL https://ollama.com/install.sh | sh

# Start Ollama
ollama serve

Windows

1. Download the installer from Ollama's official website
2. Run the installer
3. Ollama will start automatically as a system service

Verifying Installation

ollama --version

Quick Start

1. Install the extension
2. Click "Login with GitHub" - the browser will open for secure authentication
3. Choose your AI provider:
   • Local AI (Recommended):
     1. Install Ollama following the steps above
     2. Start Ollama service if not running
     3. The extension will automatically detect Ollama
     4. Select a model when prompted (llama2, mistral, phi, or neural-chat)
   • Cloud AI: Add OpenAI API key
4. That's it! StayGreen starts monitoring automatically

Commands

• StayGreen: Login - Authenticate with GitHub (starts monitoring automatically)
• StayGreen: Stop - Pause monitoring
• StayGreen: Start - Resume monitoring
• StayGreen: Change Ollama Model - Switch between available Ollama models
• StayGreen: Set OpenAI API Key - Configure cloud AI commit messages

Settings

• staygreen.schedule: Cron schedule for checking changes (default: "_/30 _ * * *" - every 1 hour)
• staygreen.ollamaModel: Selected Ollama model for local AI processing
  • llama2: General purpose model (default)
  • mistral: Fast and accurate for technical analysis
  • phi: Small but powerful model for code
  • neural-chat: Optimized for conversation
• staygreen.openaiApiKey: OpenAI API Key for cloud AI processing (optional)

AI Processing Options

Local Processing with Ollama

• No API keys needed
• Full privacy - all processing done locally
• Choose from multiple optimized models
• Perfect for sensitive codebases

Cloud Processing with GPT-4

• No local model downloads needed
• Faster processing
• More consistent results
• Better for resource-constrained systems

Troubleshooting Ollama

Common Issues

1. "Cannot connect to Ollama" error

   • Ensure Ollama is running with ollama serve
   • Check if the service is running on port 11434
   • Try restarting Ollama

2. Model download issues

   • Ensure you have stable internet connection
   • Check available disk space
   • Try downloading using command line: ollama pull modelname

3. Slow responses

   • Consider using a smaller model like 'phi'
   • Check system resources usage
   • Ensure no other intensive processes are running

4. Windows-specific issues

   • Verify Windows Defender isn't blocking Ollama
   • Run Ollama as administrator
   • Check Windows Services to ensure Ollama service is running

5. Model not found

   • Try manually pulling the model: ollama pull modelname
   • Verify model name is correct
   • Check Ollama's model list: ollama list

For more detailed troubleshooting, visit Ollama's documentation.

Security

• Uses secure GitHub OAuth flow
• No token storage needed
• Private repository for tracking
• Local changes only tracked when VS Code is running
• Option for complete local AI processing

How it Works

1. Every 1 hour, StayGreen checks for code changes
2. If changes are found:
   • AI analyzes the changes (using Ollama locally or OpenAI in cloud)
   • Creates a daily organized summary
   • Updates your contribution graph
3. All summaries are stored in date-organized files
4. Your coding activity is properly reflected in GitHub

License

MIT

Support

Found a bug or have a suggestion? Directly message me on Twitter or Linkedin

Made with ❤️ by hritik2002