Git Paid 💰 - AI Time Tracking

"Get paid for your git commits!" (but we track all your edits)

Privacy-first AI-generated work descriptions for developer time tracking. Track your coding sessions and generate invoice-ready summaries grounded only in observable signals.

Features

• 💰 Activity Tracking: Automatically tracks file opens, edits, saves, and focus changes
• 📊 Work Block Segmentation: Groups activity into meaningful "work blocks" separated by idle gaps
• 🔀 Git Integration: Collects branch names, commit messages, and diff statistics
• 🤖 AI Descriptions: Generates human-readable descriptions using VS Code's Copilot or external providers
• 🔒 Privacy First: Never sends source code off-machine; uses only metadata
• 💾 Persistent Storage: Activity data persists across VS Code sessions
• ☁️ Cloud Sync: Backup worklogs to GitHub Gists
• 📅 Monthly Exports: Generate invoice-ready monthly reports with earnings calculation
• 💵 Earnings Tracking: Set your hourly rate and see calculated earnings

Privacy & Security

This extension is designed with privacy as a core principle:

• ✅ No source code is ever sent to AI providers
• ✅ Only metadata is used: file names, paths (optionally redacted), languages, time spent
• ✅ Git context is limited to: branch names, commit messages, diff stats (lines added/removed)
• ✅ API keys are stored in VS Code's SecretStorage (encrypted)
• ✅ Local-first option: Use Ollama for fully offline operation
• ✅ Configurable redaction: Choose between full paths, basenames only, or hashed paths

What is sent to AI providers

The following metadata may be sent (depending on your settings):

{
  "workspace": "my-project",
  "duration": { "minutes": 45, "formatted": "45m" },
  "files": [
    { "name": "index.ts", "language": "typescript", "timeSpentMinutes": 20 },
    { "name": "utils.ts", "language": "typescript", "timeSpentMinutes": 15 }
  ],
  "git": {
    "branch": "feature/ABC-123-new-feature",
    "commits": [{ "message": "Add new utility functions" }],
    "diffStats": { "filesChanged": 3, "linesAdded": 50, "linesRemoved": 10 },
    "ticketIds": ["ABC-123"]
  },
  "topFiles": ["index.ts", "utils.ts"]
}

What is NOT sent:

• File contents
• Code snippets
• Full file paths (when redaction is enabled)
• Any other sensitive data

Installation

From Source (Development)

1. Clone the repository and navigate to the aw-worklog directory
2. Install dependencies:

   npm install
   

3. Compile the extension:

   npm run compile
   

4. Press F5 in VS Code to launch the Extension Development Host

From VSIX

1. Download or build the .vsix file
2. Install via VS Code: Extensions → ... → Install from VSIX

Usage

Commands

Open the Command Palette (Ctrl+Shift+P / Cmd+Shift+P) and search for:

Workflow

1. Work normally - The extension tracks your activity in the background
2. Generate descriptions - Run one of the generate commands when ready
3. Review - Use the interactive review to accept, edit, or regenerate descriptions
4. Save - Worklogs are saved to .git-paid/YYYY-MM-DD.md and .git-paid/YYYY-MM-DD.json
5. Export - Generate monthly reports for invoicing

Status Bar

The status bar shows:

• 💰 123 - Number of events recorded today
• 💰 Paused - Tracking is paused

Click to toggle pause.

Configuration

Access settings via File → Preferences → Settings → search for git-paid.

AI Providers

VS Code Language Model API / GitHub Copilot (Recommended)

Uses VS Code's built-in Language Model API with GitHub Copilot.

• ✅ No additional setup required if you have Copilot
• ✅ Uses your existing Copilot subscription
• ⚠️ Requires VS Code 1.85+ with Language Model API support

Ollama (Local)

Run AI models locally for complete privacy.

1. Install Ollama
2. Pull a model: ollama pull llama3.2
3. Set provider to ollama in settings
4. Configure endpoint and model name if needed

OpenAI

1. Get an API key from OpenAI
2. Run Git Paid: Set API key and select OpenAI
3. Set provider to openai in settings

Anthropic

1. Get an API key from Anthropic Console
2. Run Git Paid: Set API key and select Anthropic
3. Set provider to anthropic in settings

Output Format

Markdown (.worklog/YYYY-MM-DD.md)

# Worklog: 2024-01-15

**Workspace:** my-project
**Total Focused Time:** 3h 45m

## Summary

### Development of API endpoints and unit tests

- Implemented new REST endpoints for user management
- Added unit tests for authentication service
- Fixed bug in data validation logic

## Work Sessions

### 09:00 - 11:30 (2h 30m)

**Implemented user authentication endpoints**

- Created login and registration endpoints
- Added JWT token generation
- Wrote integration tests

**Files:**
- `auth.controller.ts` (typescript) - 1h 20m (45 edits)
- `auth.service.ts` (typescript) - 50m (32 edits)

**Git:**
- Branch: `feature/USER-123-auth`
- Tickets: USER-123
- Changes: 3 files, +150/-20 lines

---

JSON (.worklog/YYYY-MM-DD.json)

Structured JSON for programmatic access and integration with time tracking tools.

Development

Project Structure

aw-worklog/
├── src/
│   ├── extension.ts      # Main entry point
│   ├── types.ts          # TypeScript interfaces
│   ├── activityTracker.ts # Event tracking
│   ├── segmenter.ts      # Work block segmentation
│   ├── gitCollector.ts   # Git metadata collection
│   ├── aiProvider.ts     # AI provider implementations
│   ├── worklogManager.ts # Worklog file management
│   └── reviewUI.ts       # Interactive review UI
├── test/
│   └── suite/            # Test files
├── package.json
├── tsconfig.json
└── README.md

Building

npm install
npm run compile

Testing

npm test

Running in Development

1. Open the aw-worklog folder in VS Code
2. Press F5 to launch the Extension Development Host
3. Test the commands in the new VS Code window

Troubleshooting

No events recorded

• Check that files you're editing aren't in the excludePatterns
• Verify tracking isn't paused (check status bar)
• Some file schemes (like untitled:) are excluded by default

AI generation fails

• VS Code LM: Ensure you have Copilot or another Language Model extension installed
• OpenAI/Anthropic: Verify your API key is set correctly
• Ollama: Ensure the Ollama server is running and the model is pulled

Worklogs not saving

• Ensure you have write permissions to the workspace
• Check that the worklog path setting is valid

Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

License

MIT

Acknowledgments

Inspired by ActivityWatch and the aw-watcher-vscode extension.