Educational reference. All trademarks belong to their respective owners: Anthropic (Claude), Anysphere (Cursor), Microsoft/GitHub (Copilot).Educational reference for Claude, Cursor, and Copilot.
Claude
Slash commands, skills, subagents, and hooks for Claude Code (terminal CLI). Surface tags mark entries that also work on Desktop, IDE, Chrome, mobile, or integrations.
Default surface: Most entries run in the Claude Code terminal. Additional tags show when the same command also targets Desktop, IDE, Chrome, or other integrations.
Click any entry to expand a step-by-step run guide — where to access it, how to run it, and what to expect.
Core Commands
Session & Context
Review & Quality
Configuration
Account & Feedback
Integrations & Surfaces
Plugins & Skills
Workflows & Automation
Experience
About Claude Code commands, skills, agents, and hooks
Claude Code is Anthropic's terminal-native coding agent. Instead of copying answers out of a chat window, you run it inside your project and drive it with slash commands, reusable skills, subagents, and lifecycle hooks.
This page collects the slash commands and extension points that most Claude Code users reach for day to day, each with a short description of what it does and a copy-paste example. The goal is to save you from digging through release notes and scattered docs when you just need to remember how a command is invoked or what it expects.
Slash commands cover the interactive basics — starting and clearing context, reviewing changes, and running focused edits. Skills package repeatable instructions that Claude can discover and apply automatically. Subagents let you hand a scoped task to a separate context so your main session stays clean, and hooks fire on events such as a permission prompt or an idle teammate so you can automate guardrails.
What you'll find on this page
Slash commands: Interactive terminal commands for context, review, and edits.
Skills: Reusable instruction packages Claude can auto-invoke by trigger.
Subagents: Scoped helpers that run in a separate context for deep work.
Hooks: Event-driven automation for permissions, lifecycle, and safety.
Frequently asked questions
Where do Claude Code commands run?
Most entries run in the Claude Code terminal. Some also work in the Desktop app, IDE extension, or Chrome extension — the surface tags on each card show where a command is available.
Are these official Anthropic commands?
Command names come from official Claude documentation and community usage, but this is an independent, community-maintained reference. Always confirm against Anthropic's docs before relying on a command in production.
How often is the Claude catalog updated?
Entries are refreshed as Anthropic ships changes. Check the What's new page for recently added commands, or subscribe on the feedback page to get an email when the catalog grows.