Git extension that tracks every line of AI-generated code across all coding agents (Cursor, Copilot, Claude) with vendor-agnostic attribution that survives rebases, merges, and squashes

Remember when we could tell human-written code from AI-generated code at a glance? Those days are over. As teams adopt multiple AI coding agents, tracking what’s AI-generated becomes impossible—until now. Git AI solves the “AI code accountability” problem by automatically tagging every AI-generated line with its source agent, model, and original prompt, creating an audit trail that persists through your entire Git workflow.

What sets this apart is its cross-platform, vendor-agnostic approach. Works with Cursor, GitHub Copilot, Claude Dev, and more simultaneously. The Rust implementation adds negligible overhead (<100ms even in massive repos like Chromium), and the Git Notes-based storage means your AI attribution data survives rebases, squash merges, and cherry-picks—operations that typically destroy code provenance. The git ai blame command becomes your new best friend for code reviews.

Perfect for teams juggling multiple AI tools or organizations needing compliance around AI code usage. One-line install, zero per-repo configuration, immediate tracking of all supported agents. With 947 stars and growing, this is becoming the de facto standard for AI code attribution in professional environments.

⭐ Stars: 947
💻 Language: Rust
🔗 Repository: git-ai-project/git-ai