Git for your AI agent’s actions. Undo, trace, and control every step. re_gent shows what your coding agent changed, which prompt caused it, and lets you roll back agent work across files and sessions.
Loading comments…
Achievement
Project Info
Product Keywords
Re_gent is version control built specifically for AI agent activity. While Git tracks file changes, Re_gent tracks the conversations and prompts that produced those changes. It captures every action your coding agent takes, shows you exactly which prompt caused each modification, and lets you undo, blame, checkout, or replay agent work across multiple sessions. Think of it as Git for your AI agent's actions — a complete audit trail that keeps you in control.
rgt logRe_gent shows you the full conversation that led to each change. Instead of a vague "fix bug" commit message, you see the actual prompt that triggered the modification. Every agent action is recorded with its context, making it easy to understand why something changed.
rgt blameEvery line of code is traced back to its exact prompt. If three different agent sessions touched the same file, Re_gent shows you which prompt from which session caused each change. This gives you a complete audit trail across multiple agent interactions.
rgt sessionsEach agent conversation gets its own branch, allowing multiple agents to work simultaneously without collisions. You can review and merge their work just like you would with human developers, keeping everything organized and conflict-free.
Even when AI tools like Claude compact or delete messages, Re_gent preserves the complete record in content-addressed storage. You never lose the audit trail, no matter what the agent does to clean up its own history.
"Git tracks files. Re_gent tracks the agent conversations that produced those files."
This distinction is the core of Re_gent's value. Traditional version control treats code changes as isolated events, but modern AI development is conversational. When an agent breaks something, Re_gent doesn't just show you the broken code — it shows you the exact prompt that caused the breakage and lets you rewind both the code and the conversation. It's infrastructure designed for a world where AI agents write most of the code.
You use AI coding assistants like Claude Code, Codex, or OpenCode and want more control over what your agents do. If you've ever wondered "what did my agent actually change?" or needed to undo a specific agent action without losing unrelated work, Re_gent gives you that power. It's especially valuable for teams running multiple agents in parallel, or anyone who wants a reliable audit trail for AI-generated code. Best of all, it's free forever under the Apache-2.0 license and open-source on GitHub.
Other tools you might consider
Runsight is a YAML-first workflow engine designed specifically for AI agents , enabling developers to design, commit, run, and evaluate agent workflows with Git-native version control. Every workflow is stored as a YAML file in your repository, allowing you to branch, review, and merge changes just like any other code. The platform offers real-time cost tracking per run with hard budget caps to prevent overspending, along with a built-in evaluation framework for assertions and regression testing. "Ship agents like you ship code." Feature Benefit Canvas + YAML Editor Dual visual and code views Per-run Cost Tracking Monitor spending to the cent Git Integration Version control for workflows It's completely self-hosted, runs on your machine with your API keys, and is 100% open source under Apache 2.0 license.
GenAI is like a race car without brakes. It accelerates fast — you describe something, and large blocks of code appear instantly. But AI agents silently break things. They remove logic. Relax constraints. Introduce expensive cloud calls. Leak credentials. Change behavior without telling you. git-lrc is your braking system. It hooks into git commit and runs an AI review on every diff before it lands.
Axel helps you run AI agents and keep them fed. Queue up work, dispatch to the right agent, and approve or deny actions from one inbox. It's native macOS, keyboard-driven, and works with Claude, Codex, OpenCode, and Antigravity out of the box. We hope it helps you ship faster 🚀
Whats 1Code? An app to run your Claude Code agents in parallel that works on Mac and Web. On Mac - run locally, with or without worktrees. On Web - run in remote sandboxes with live previews of your app, mobile included, so you can check on agents from anywhere. Running multiple Claude Codes in parallel dramatically sped up how we build features.
Maker
neon_dev
Compare with
Alternatives
Loading comments…