Code Detective 🕵️♂️
Git blame tells you WHO. Code Detective tells you WHY.
An AI-powered code archaeology tool powered by Gemini 3. Uncover the hidden history, context, and expertise behind every line of code.
🚀 Why Code Detective?
Every developer has experienced the "Legacy Code Fear": staring at a block of complex, undocumented code and wondering, "Why does this exist? Who wrote this? If I touch it, will everything break?"
Code Detective solves this by acting as an AI archaeologist. It doesn't just read the code; it reads the history of the code.
- Understand Intent: Discover the business logic and edge cases behind "weird" code.
- Identify Experts: Instantly know who to ask for help with the Expertise Map.
- Measure Health: Get a concrete Tech Debt Score and refactoring plan.
- Onboard Faster: Generate documentation for any part of the codebase in seconds.
✨ Features
🔍 AI Investigation
Right-click any code to launch a deep investigation.
- Contextual Analysis: Traces git history to find the original PRs and issues.
- Deep Dive: Autonomous agents cross-reference multiple files to build a complete picture.
- Thinking Process: Watch Gemini 3's "High Thinking" mode reason through complex history in real-time.
📜 Time Travel Timeline
A beautiful, interactive visualization of your code's evolution.
- Interactive History: Scroll through the life of a file.
- Diff Analysis: Click any commit to see exactly what changed.
- Searchable: meaningful search across all commit messages and authors.
📊 Tech Debt Score
Gamify your code quality.
- AI Grading: Get an A-F grade for your project.
- Factor Analysis: Break down debt by Freshness, Clarity, Complexity, and Consistency.
- Actionable Advice: AI suggests specific steps to improve your score.
👥 Who Knows What? (Expertise Map)
Visualize team knowledge distribution.
- Expertise Analysis: Scans commit history to see who owns which modules.
- Bus Factor: Identify parts of the code only one person understands.
- Find Mentors: Know exactly who to ping for help.
📘 Instant Onboarding
Turn any folder into a handbook.
- Auto-Documentation: Generates
Overview, Architecture, and Key Files docs.
- Gotchas: AI highlights common pitfalls and patterns in the module.
⚡ Smart Utils
- Commit Message Generator: Analyses staged changes and writes Conventional Commits for you.
- Watch Mode: Background agent that warns of merge conflicts and remote updates.
🧠 Powered by Gemini 3
Code Detective utilizes the advanced capabilities of the Google Gemini 3 model:
- Thinking Level HIGH: We enable Gemini's deep reasoning mode to deduce intent from sparse commit messages.
- 1M Token Context: We feed massive amounts of git history and file context into the model for comprehensive analysis.
- Multi-Agent System:
- Lead Detective: Orchestrates the investigation.
- Historian: Mines git data.
- Archivist: Analyzes external context.
🛠️ Installation & Setup
- Install the Extension from the VS Code Marketplace.
- Open a Valid Git Repository in VS Code.
- Configure API Key:
- Get a Gemini API Key from Google AI Studio.
- Run command
Code Detective: Set API Key OR set GEMINI_API_KEY in your environment.
Configuration Settings
| Setting |
Description |
Default |
codeDetective.geminiApiKey |
Your Google Gemini API Key |
null |
codeDetective.thinkingLevel |
AI Reasoning depth (low, medium, high) |
high |
codeDetective.githubToken |
(Optional) GitHub Token for PR analysis |
null |
🎮 Usage
Start an Investigation
- Select any code in your editor.
- Right-click and choose "Investigate Code".
- The Code Detective panel will open with your results.
View Timeline
- Open the Command Palette (
Cmd+Shift+P).
- Run "Code Detective: View Code Timeline".
Analyze Tech Debt
- Click the Code Detective icon in the Activity Bar (Telescope).
- Click "Tech Debt Score" in the sidebar.
Generate Onboarding Docs
- Click "Onboarding Docs" in the sidebar.
- Select the folder you want to document.
🤝 Contributing
We welcome contributions! Please see our GitHub Repository for details.
- Fork the repo.
- Create your feature branch (
git checkout -b feature/AmazingFeature).
- Commit your changes (
git commit -m 'Add some AmazingFeature').
- Push to the branch (
git push origin feature/AmazingFeature).
- Open a Pull Request.
📄 License
Distributed under the MIT License. See LICENSE for more information.
Built with ❤️ for the Google DeepMind Gemini 3 Hackathon.
