English | 简体中文

WTF Commit ✨

Links: GitHub · VS Marketplace · Open VSX · Website

One shortcut. Your commit message is written, committed, and pushed — in seconds.

WTF Commit reads your git diff, writes a clear Conventional Commit (e.g. feat: add dark mode), fills the Source Control box, and commits in the same keystroke — push too, once you enable Auto Push. No pausing to phrase a message, no boilerplate.

Flow: git diff → ✨ shortcut → message → commit → (optional) push

Already installed? Skip ahead: ① Configure AI (~1 min) → ② Daily use.

Install

VSIX (optional / offline): download from GitHub Releases or the marketplaces above, then Extensions → ⋯ → Install from VSIX…, or drag the .vsix onto the Installed list. Manual VSIX installs do not auto-update — prefer the marketplace install when you can.

① Configure AI (once)

Defaults are enough: Provider = DeepSeek. You only need a key.

1. Install — search WTF Commit in your editor’s Extensions view (VS Code or Cursor).
2. Set API Key — Command Palette → WTF Commit: Set API Key, then:
   • The first entry is your current provider (it shows whether a key is already stored); pick Choose another provider… to switch.
   • First-time setup shows a small wizard: press Enter on Paste API Key now, or pick Open {provider} key page — the wizard stays open while the browser opens, so you can come back and paste once you have a key.
   • Pasting is masked with live paste-error checks; on success the status bar echoes the masked key (e.g. sk-ab••••wxyz) so you can verify what was stored.
   • A key saved for another provider asks whether to Switch Provider.
3. Need a key? DeepSeek API keys (linked from the extension too).

Want another provider later? In Set API Key choose Choose another provider…, or change Provider in Settings. Leave Model empty for built-in defaults.

② Daily use (every commit)

What the shortcut does: prefers staged changes → if nothing is staged and Auto Commit is on, stages your working tree → generates a Conventional Commit → commits (default) → does not push (Auto Push off).

1. Make your code changes (stage when you can).
2. Press the generate shortcut — the message streams in, then commits by default.
3. Push yourself when ready — Auto Push is off by default.

Default shortcut: Cmd+Alt+G (Mac) / Ctrl+Alt+G (Windows/Linux).

Make it yours — the binding is fully customizable. A popular Cursor-style habit is a double press of Cmd+G / Ctrl+G (chord):

1. Open Keyboard Shortcuts (Cmd+K Cmd+S / Ctrl+K Ctrl+S).
2. Search WTF Commit: Generate.
3. Double-click the keybinding → press Cmd+G twice (or Ctrl+G twice on Windows/Linux) → Enter.

One-keystroke power flow (optional): Settings → enable Auto Push. Leave Confirm Before Push on until you trust it; turn that confirm off only when you want a zero-dialog generate → commit → push flow.

Other triggers: ✨ on the Source Control title bar, or Command Palette → WTF Commit: Generate.

Prefer review-only? Turn Auto Commit off — the message stays in Source Control for you to edit and commit manually.

🆕 Latest (v1.20.1)

• Cheaper OpenAI default: OpenAI now uses gpt-6-luna with reasoning disabled for short commit messages.
• OpenCode Go model ID: Its default now uses the documented deepseek-v4.1-flash ID.

See CHANGELOG for full history.

🚀 Features

• Conventional Commits — feat / fix / docs / … with optional local format fix + automatic AI repair.
• Smart diffing — Prefers staged changes; with Auto Commit, stages the working tree when nothing is staged.
• Intent-aware — Text already in the SCM input is used as a generation hint (no extra prompt UI).
• Streaming preview — Watch the message appear live while the model runs.
• Auto Commit & Push — Optional one-keystroke pipeline with a push confirmation.
• Multi-language messages — English, 简体/繁体中文, Japanese, Classical Chinese, or Custom.
• Bring your own endpoint — Built-in providers plus Custom (Ollama, proxies, etc.).
• Keyboard shortcut — Default Cmd+Alt+G / Ctrl+Alt+G; rebind freely (e.g. double Cmd+G).

🛠️ Advanced Tutorial

1. Plugin Settings

Open VS Code Settings (Cmd+,) and search for WTF Commit. Settings are split into WTF Commit (basics) and WTF Commit › Advanced.

Basics — what most people need:

Advanced — Custom Base URL/Model, Provider Overrides, ignore paths, status bar toggle.

Extension UI language follows VS Code (vscode.env.language): zh* → 中文, otherwise English. Power-user overrides live in settings.json only: wtfCommit.prompt, and when Commit Message Language is Custom, wtfCommit.customCommitMessageLanguage.

2. Custom Model & Endpoints

You can use any OpenAI-compatible model (like local models via Ollama):

1. In Settings, set Provider to Custom.
2. Under WTF Commit › Advanced, enter the Base URL (e.g., http://localhost:11434/v1).
3. Enter the Model name (e.g., llama3).

To override a built-in provider's endpoint/model, use Provider Overrides (not the global Base URL / Model fields).

3. Custom Commit Language

If you want the AI to use a specific language (e.g., French, Cantonese, or Emoji-only):

1. Set Commit Message Language to Custom.
2. In settings.json, set the target language, for example:

{
  "wtfCommit.commitMessageLanguage": "Custom",
  "wtfCommit.customCommitMessageLanguage": "Emoji only"
}

ℹ️ Supported Providers & Models

How defaults work (three separate concepts):

Provider defaults — when Base URL and Model are left empty:

Get API Keys

Other OpenAI-compatible APIs (MiMo, GLM, Z.AI, NVIDIA NIM, proxies): set Provider to Custom and fill Base URL + Model under Advanced — or route through OpenRouter.

Choosing a model for commit messages

Generating a commit message is a lightweight task — you don't need a frontier model. Pick based on your use case, cost preference, and whether you already have an API key:

Recommended path:

1. 🎯 Out-of-the-box / Pay-as-you-go (Recommended) 👉 DeepSeek (deepseek-flash)
   • Default Provider; leave Model empty.
   • True pay-as-you-go with no subscription barrier. Fast, high quality, and thinking mode is disabled automatically for sub-second responses.
2. 🎁 Existing Coding Sub / Global Access (Recommended) 👉 OpenCode Go (deepseek-v4.1-flash)
   • Recommended if you already have (or want) an OpenCode Go subscription: set Provider to OpenCode Go, paste the Go API key.
   • Low-latency global endpoints; commits draw from your Go quota. DeepSeek V4.1 Flash currently has a limited-time 4× usage credit.
3. 🆓 Zero Cost / Free Tier 👉 Gemini (gemini-3.5-flash-lite)
   • Google AI Studio provides a generous free tier (15 RPM / 1M tokens/day). Uses native Interactions API with minimal thinking for instant responses.
4. 🔑 OpenAI Key Holders 👉 OpenAI (gpt-6-luna)
   • Low-cost model for focused tasks, configured without reasoning for short commit messages.
5. 🧪 Zero-Setup Trial / Self-Hosted 👉 OpenRouter (openrouter/free trial) or Custom (local Ollama, MiMo, GLM, enterprise proxies).

DeepSeek vs OpenCode Go (same model): DeepSeek uses deepseek-flash; Go documents deepseek-v4.1-flash. Both select V4.1 Flash and list the same rates (off-peak: $0.15 input / $0.60 output; peak: $0.30 input / $1.20 output per 1M tokens). Go includes a $10/mo subscription quota. DeepSeek V4.1 Flash currently has a limited-time 4× usage credit on Go (see Go pricing; promo terms can change). For light use without an existing subscription, compare pay-as-you-go prices; if you already pay for Go, commits can use its quota.

Pricing comparison (USD per 1M tokens, cache-miss input; sources linked below):

• DeepSeek & OpenCode Go peak rates apply 01:00–04:00 and 06:00–10:00 UTC (Mon–Fri); all other times are off-peak (50% discount).

† Rough estimate for ~5K input + 150 output tokens (typical diff + commit message), no prompt cache. Actual cost depends on diff size, tokenization, and model output length.

‡ OpenCode Go publishes the same Flash list rates as DeepSeek to meter subscription quota (Go pricing). DeepSeek V4.1 Flash currently has a limited-time 4× usage credit; within the included credit, marginal cost for commits is effectively ~$0 until you hit Go limits.

Official pricing pages: DeepSeek · OpenCode Go · Gemini · OpenAI · OpenRouter

OpenRouter provider default is openrouter/free.

Gemini uses Google's native Interactions REST API (/v1beta/interactions), authenticates with the x-goog-api-key header, and uses the minimal thinking level to reduce latency for commit-message generation.

Upgrading from an older build that used MiMo / GLM / Z.AI / NVIDIA NIM: the extension migrates that Provider setting to Custom and fills Base URL / Model automatically.

[!IMPORTANT] Claude Support: Native Claude format is not supported yet. Please use a proxy service that provides an OpenAI-compatible endpoint.

🕹️ Other entry points

• Source Control title bar ✨ · Command Palette WTF Commit: Generate · rebindable shortcut (see Daily use).
• Rotate keys anytime with WTF Commit: Set API Key.

💬 Feedback

Using WTF Commit and finding it useful? A GitHub star, an Open VSX review, or a short issue (bugs or ideas) all help more people discover it — and help us keep improving.

📄 License

MIT License.