Autogit — AI skill for Claude Code
Auto commit & push for AI coding agents, with a safety level you pick.
How to install Autogit
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open davidondrej/autogit and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Autogit does
Auto commit & push for AI coding agents, with a safety level you pick.
Alternatives in AI
- Oil UI — Push AI UI design to its limits: explore distinct directions, compare them side by side, and refine against re 861 ★
- Agentpet — A desktop pet for macOS & Windows that monitors your AI coding agents (Claude Code, Codex, Cursor, Gemini 334 ★
- Crux — Taking the pi coding agent to Claude Code-level performance: 0.539 → 0.773 pass@1 on Terminal-Bench 2.1 with t 187 ★
README
autogit
Your AI coding agent writes the code. **autogit ships it.**
When your agent finishes a turn, autogit stages, commits, and pushes — automatically.
Quick start
# 1. Install (once per machine)
npm install -g @davidondrej/autogit
autogit setup
# 2. Enable it per repo
cd your-project
autogit on
Done. Every agent turn now ends with: **stage → secrets scan → commit → push.**
From source instead: `git clone https://github.com/davidondrej/autogit && cd autogit && npm link`
Supported agents
| Agent | After autogit setup |
|---|---|
| Claude Code | works immediately |
| Cursor | works immediately — interactive local + worktree agents, including repos entered or created after the session starts (agent -p has no stop hook) |
| Pi | works immediately |
| Codex | one-time approval: restart open sessions, then run /hooks in codex and trust autogit (needs ≥ 0.124) — covers the CLI, the Codex desktop app, and the IDE extension |
Hooks fire for interactive local agent sessions. Cursor `agent -p`, Cursor cloud agents (no local user hooks), Codex cloud tasks, and `codex exec` are not covered — upstream limitations. Codex re-asks for `/hooks` trust whenever autogit updates its hook entries.
Commands
autogit setup Wire up agent hooks (once per machine)
autogit on Enable auto-push in this repo
autogit on --agent Agent mode: an LLM reviews each turn before it ships
autogit off Disable auto-push in this repo
autogit ship Stage, scan, commit, push (what the hooks run)
autogit undo Take back the last autogit commit, local + remote
autogit status Show hooks + repo state
autogit update Update autogit to the latest version from npm
autogit --version Print the installed version (-v)
**Commit messages**: `autogit ship -m "message"` uses your message. Without `-m`, the subject is the prompt you gave your agent that turn
Related Skills
Wechat Auto Publisher
WeChat Official Account automation: pick articles from your Obsidian/Markdown vault, AI-rewrite & push to draf
Aigate AI Git Workflow Guard CLI
Pre-push Git safety CLI for AI-assisted coding: catch risky changes, possible secrets, weak PR readiness, and
AI Push
Commit and push all changes in ~/.claude/ (the AI brain) to GitHub, regenerate the neural connectome, and sync
Declaude
Remove Claude/AI attribution from a GitHub repo in one command clean commit history, force-push, and refresh t
Agents In A Chest.Nvim
Neovim plugin to run LLM coding agents in isolated Git work trees with auto‑commit, tabbed sessions, and safe
CommitLM
Teach your LLM coding agents what just happened: Auto docs and commit message generation, with LLM
Related Agents
Scala Coder
Implements a single coding task end-to-end in an assigned git worktree for this Scala project. Owns the FULL l
Wave Babysitter
Babysits a wave of background agents: watches their output files, auto-resumes API/limit deaths ONCE per agent
Opensource Sanitizer
Pre-public-push safety net. Scans a diff or working tree for leaked secrets, PII, internal references, interna