#180180th of 236 in Coding agents and AI editors

opencommit

CLI that writes git commit messages from staged diffs using an LLM

Stars
7.5k
License
MIT
Last commit
Nov 2026
Last release
May 2023
Language
JavaScript
Stars, last 29 weeks+361
25 Apr 2026: 7,188 stars2 May 2026: 7,198 stars9 May 2026: 7,212 stars16 May 2026: 7,221 stars23 May 2026: 7,228 stars30 May 2026: 7,238 stars6 Jun 2026: 7,248 stars13 Jun 2026: 7,261 stars20 Jun 2026: 7,271 stars27 Jun 2026: 7,275 stars4 Jul 2026: 7,363 stars11 Jul 2026: 7,412 stars18 Jul 2026: 7,426 stars25 Jul 2026: 7,444 stars1 Aug 2026: 7,451 stars8 Aug 2026: 7,461 stars15 Aug 2026: 7,470 stars22 Aug 2026: 7,478 stars29 Aug 2026: 7,485 stars5 Sept 2026: 7,496 stars12 Sept 2026: 7,499 stars19 Sept 2026: 7,509 stars26 Sept 2026: 7,514 stars3 Oct 2026: 7,516 stars10 Oct 2026: 7,521 stars17 Oct 2026: 7,525 stars24 Oct 2026: 7,534 stars31 Oct 2026: 7,538 stars7 Nov 2026: 7,541 stars14 Nov 2026: 7,549 stars

7,188 on 25 Apr 20267,549 on 14 Nov 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

OpenCommit is an npm-installed CLI (`oco`) that sends your staged diff to an LLM and commits with the generated message. It supports OpenAI, Anthropic, Azure, Gemini, Ollama, llama.cpp, Flowise, DeepSeek and OpenRouter, and can run as a prepare-commit-msg hook or a GitHub Action that rewrites commit messages on push.

Who it is for: Developers who want AI-written commit messages in the terminal, hooks or CI

Strengths

  • Works with local models via Ollama or llama.cpp, so diffs need not leave your machine
  • prepare-commit-msg hook integrates with IDE source control and allows editing before commit
  • Per-repo .env and global ~/.opencommit config, plus .opencommitignore to exclude files from the prompt
  • Supports conventional-commit or commitlint prompt modules, GitMoji, and multiple output languages

Weaknesses

  • GitHub Action is marked beta and rewrites history via rebase, changing commit SHAs
  • Last tagged release is from 2023-05-21 despite recent commits
  • Default input limit is 4096 tokens, so large diffs may need tuning
  • README contradicts itself on default model: gpt-4o-mini in config, 3.5-turbo in payments

What it needs

  • no GPU
  • Needs Node.js, npm, git, LLM provider API key or local Ollama/llama.cpp server
  • Models: OpenAI, Anthropic, Azure OpenAI, Gemini, Ollama

What changed

From the release notes. No flag means the note does not say so, not that the upgrade is safe.

Also in Coding agents and AI editors

See all 236
Also in Coding agents and AI editors
RankProjectScore
1claw-codeRust CLI coding agent harness using Anthropic and OpenAI-compatible APIs194.8k starsMIT96 out of 100
2t3codeDesktop, web and mobile front end for controlling local coding agent CLIs Live demo ↗ (opens in a new tab)64.4k stars+36,149 (+128%) in 30 daysMIT96 out of 100
3codewhaleTerminal coding agent in Rust that works with 40+ model providers41.2k starsMIT94 out of 100
#4gooseRust AI agent with desktop app, CLI and API, extended through MCP56.7k stars+1,476 (+2.7%) in 30 daysApache-2.093 out of 100
#5ArchonYAML workflow engine that runs AI coding agents through fixed dev steps23.9k starsMIT93 out of 100