Cook documentation
Find your way around Cook
Cook has a terminal app and a desktop app. Pick one to get started, then use the guides for commands, workflows, and settings.
Cook CLI
Install and sign in, use the interactive terminal app, or run Cook in scripts and CI.
Get started with the CLI DesktopLet Cook
Choose a workspace, connect a model provider, chat, review changes, and use desktop shortcuts.
Desktop guideGetting started
- Getting Started
Cook has two apps. The Cook CLI runs in your terminal, and Let Cook is the desktop app for Windows, macOS, and Ubuntu. Both work with your local projects…
- Authentication
The Cook CLI supports browser login, API keys, enterprise single sign-on (SSO), and external auth providers for automation. This page covers the CLI. In L…
Using Cook
- Keyboard Shortcuts
Reference for key bindings in the Grok Build TUI. Bindings are built in and cannot currently be remapped.
- Headless Mode and Scripting
Headless mode runs Cook non-interactively from the command line. It accepts a single prompt, executes it with full tool access, and returns the result. Us…
- Agent mode (ACP) and IDE integration
Agent mode runs Cook as a long-lived server that clients talk to over ACP (JSON-RPC). Use it from IDEs, SDKs, eval harnesses, and custom apps. For a one-s…
- Let Cook Desktop
Let Cook is Cook's desktop app for Windows, macOS, and Ubuntu. It gives you a chat window for working in a local project, with a session list, model setti…
Commands
- Slash Commands
Type / in the prompt to open the command menu. It fuzzy-matches as you type, and picking a command runs it immediately.
- Terminal CLI Reference
Use cook --help for the options in your installed version. Use cook --help for a command's details. This page lists the main commands and common options.
Workflows
- Subagents and Personas
Subagents are independent child sessions that handle tasks in parallel. Each subagent has its own context window, so the main agent can delegate work (res…
- Session Management
Cook saves every conversation to disk automatically. Whether you work in the TUI, in headless mode, or over agent stdio, Cook records the exchange as a se…
- Plan Mode
Plan mode is a structured planning phase: the agent explores the codebase and designs an implementation approach before writing any code. Use it for tasks…
- Background Tasks and Monitoring
Cook runs long-lived processes without blocking the conversation. This document covers background commands, the /loop command, the monitor tool, and the s…
- Common Workflows
Use a goal for a long task, plan mode when you want to review steps before code changes, and compaction when the conversation is getting large.
Configure and extend
- Configuration
Cook reads settings from config files, environment variables, and CLI flags. This page covers the common options. The field list for config.toml, managedc…
- Theming and Appearance Customization
Cook draws all TUI colors from a central theme. You can switch themes while Cook is running, follow your operating system's light or dark appearance, and…
- MCP Servers
MCP (Model Context Protocol) servers extend Cook with external tool integrations. They let Cook interact with any service that implements the MCP standard.
- Skills
Skills are reusable prompt packages that extend Cook with task-specific instructions. They let you capture a repeatable procedure once, instead of re-expl…
- Plugins
A plugin bundles skills, slash commands, agents, hooks, and MCP servers into one installable unit. You get plugins from a marketplace, install the ones yo…
- Hooks
Hooks let you run a script or send an HTTP request at key moments in a Cook session. Use them to automate tasks, enforce safety checks, log activity, send…
- Custom Models
Cook connects to custom model endpoints for alternative providers, self-hosted models, and overriding built-in settings. This guide explains how to select…
- Project Rules (AGENTS.md)
Project rules let you configure Cook per project or directory. By placing an AGENTS.md file in your repository, you can set coding conventions, build inst…
- Cross-Session Memory
Memory lets Grok recall facts, decisions, and patterns from earlier sessions. Grok indexes the information you save and searches it automatically, so a ne…
- Sandbox Mode
Sandbox mode restricts what the agent process and its spawned commands can access on your filesystem and network using OS-level kernel primitives (Landloc…
- Permissions and safety
Control what Cook can access and do: permission modes, allow/ask/deny rules, hooks, and the optional OS-level sandbox.
- Status Line
An optional row at the bottom of the pager — above the shortcuts bar in the full screen, under the prompt's info row in minimal mode — and disabled by def…
- Configuration reference
This file ships with the CLI and is extracted to ~/.cook/docs/user-guide/26-config-reference.md on launch. It is the complete field list for config.toml,…
Reference
- Terminal Support and Troubleshooting
Cook runs as a full-screen TUI. It relies on terminal support for color, clipboard, keyboard input, mouse input, and full-screen display. Terminals, multi…
- Agent Dashboard
The Agent Dashboard lists every top-level session in this pager process — local sessions and forks — grouped by state. From one screen you can peek, reply…
- Monitoring Usage (External OpenTelemetry)
Status: alpha. The schema below is versioned (grokcode.schema.version = v1); additive changes may occur without notice, renames/removals will bump the ver…
- grok clone
grok clone fetches a Git repository into a Grove content store and mounts a projected working tree (NFS on macOS, FUSE on Linux). Each invocation reads GR…
- Prompts
Cook injects a set of model-facing prompt templates: the plan-mode reminders, the goal templates, and the built-in subagent bodies. Each one ships as the…