Getting started
Four steps from one install command to a fleet set up for your work.
Install Codewhale
One npm command installs the terminal runtime. Cargo, prebuilt archives, Docker, Nix, and China mirrors also work. All of them install published releases.
Full install guide →npm install -g codewhale codewhale doctorOpen a first session — no key needed
Starts without any API key. A short setup, then the full interface. Look around in Plan mode, which is read-only. Model replies need a provider — that's the next step.
Learn the product nouns first →codewhaleConnect a provider
Use a hosted API key, a gateway, or a local runtime with no key (Ollama, vLLM, SGLang). You pick the provider and the model. A model name never switches the provider for you.
Providers and models →codewhale auth set --provider deepseekSet up your ideal fleet
Add each provider you use with one auth set (local runtimes need none). Then run /fleet setup. It goes one member at a time — a semantic role, a model from any configured provider, a thinking tier, then an exact identity/route review — and saves the roster profile for this repo or for every repo on this machine. Runtime permissions stay separate. A single task needs none of this.
Fleet and Workflow docs →/fleet setup codewhale fleet status
Watch a real session
A recording of a real session will go here. There is no recording yet, so nothing is shown.
There is no recording yet. When there is, it goes here with captions, a transcript, and an optional GIF download.
Where next
Hooks
Run your own commands before and after tool calls, at turn end, and on session events, with per-project trust rules.
Modes and permissions
Plan / Work / Operate and Ask / Auto-Review / Full Access: what each one allows.
Documentation hub
Every topic, searchable. Each page links to its source document in the repository.
Source documents: docs/GUIDE.md, docs/KEYBINDINGS.md · Step copy lives in web/lib/content/getting-started.ts; update docs-map.ts when changing.