AI-Native BinaryConfigure your AI provider in ~/.gitaiflow/config.envView AI-Native Binary configuration

Install the gitaiflow binary

Install gitaiflow as one native binary for macOS, Linux and Windows — no Python or package manager needed.

AI commit messages, changelogs and release notes from your terminal, scripts and CI.

Install

# latest

$curl -fsSL https://install.djangoplay.org/gitaiflow | bash

# a specific version

$curl -fsSL https://install.djangoplay.org/gitaiflow | bash -s -- v1.1.0

2
Verify the installation
```
$gitaiflow --version
```

Supported platforms and install details

Platforms: macOS (Apple Silicon), Linux (x86_64 and arm64), and Windows (x86_64). There is no prebuilt binary for Intel Macs yet — build from source instead.
macOS & Linux install to ~/.local/bin/gitaiflow with no sudo; if that folder is not on your PATH, the installer prints the line to add. Windows installs to %LOCALAPPDATA%\Programs\gitaiflow and adds it to your user PATH. Set GITAIFLOW_INSTALL_DIR to change the macOS & Linux location.
Every installer verifies a SHA-256 checksum before installing, and a failed check aborts the install.
Re-run the installer to upgrade in place; gitaiflow --uninstall removes the binary and local state files.

AI-Native Binary configuration

gitaiflow needs four settings before its first run: AI_PROVIDER, AI_API_KEY, AI_BASE_URL, and AI_MODEL. The installer creates an empty ~/.gitaiflow/config.env for exactly this — fill it in once and you are set. Any OpenAI-compatible provider works, from a free cloud model to Ollama on your own machine.

Settings are read from real environment variables first, then a project-local .env, then ~/.gitaiflow/config.env.

Example: Gemini (free tier)

AI_PROVIDER=gemini
AI_API_KEY=<your-key>
AI_BASE_URL=https://generativelanguage.googleapis.com/v1beta/openai
AI_MODEL=gemini-flash-lite-latest

Variables

AI_PROVIDER

Default: Required
Notes: Free-form provider label (gemini, ollama, openai, custom, …); only affects display and usage-log grouping.

AI_API_KEY

Default: Required
Notes: Needed for every provider, including local ones that ignore its value — use any non-empty placeholder for Ollama.

AI_BASE_URL

Default: Required
Notes: Full OpenAI-compatible chat-completions base URL.

AI_MODEL

Default: Required
Notes: Model name/id exactly as the configured provider expects it, e.g. gemini-flash-lite-latest, llama3.2:3b, gpt-4.1-mini.

Name	Default	Notes
AI_PROVIDER	Required	Free-form provider label (gemini, ollama, openai, custom, …); only affects display and usage-log grouping.
AI_API_KEY	Required	Needed for every provider, including local ones that ignore its value — use any non-empty placeholder for Ollama.
AI_BASE_URL	Required	Full OpenAI-compatible chat-completions base URL.
AI_MODEL	Required	Model name/id exactly as the configured provider expects it, e.g. gemini-flash-lite-latest, llama3.2:3b, gpt-4.1-mini.

Prefer to stay local? Ollama (AI_API_KEY still needs some non-empty value)

AI_PROVIDER=ollama
AI_API_KEY=not-needed
AI_BASE_URL=http://127.0.0.1:11434/v1
AI_MODEL=llama3.2:3b

Run it for the first time

In any Git repository with uncommitted changes, run the command below. gitaiflow reads your diff and prints a ready-to-use commit message, and saves a structured summary you can reuse for changelogs and release notes.

Summarize your changes

$gitaiflow --change-summary

Go deeper

Command reference — every flag Security and release verification Installation guide (docs)Full configuration reference (docs)Release signing and checksum verification (docs)Troubleshooting (docs)

DjangoPlay · gitaiflow