Slash commands are commands you type in Claude Code by beginning your message with a forward slash (/). They include built-in commands that control the session (for example clearing the conversation, compacting context, viewing help, or switching the model), bundled skills shipped with Claude Code, and custom commands you create to reuse your own prompts. See the official Claude Code docs for the full, current list, since the built-ins evolve across releases.
Custom slash commands are Markdown files, and they have been merged with Agent Skills: a file like .claude/commands/deploy.md and a skill at .claude/skills/deploy/SKILL.md both create a /deploy command and work the same way, so existing command files keep working. A file in your project becomes a project-scoped command you can share with your team, while one in your personal user-level directory is available across all your projects, and the command name is derived from the file (or skill) name. Commands can accept arguments and use frontmatter to configure behavior; refer to the docs for exact directory paths, argument placeholders, and frontmatter keys.
Because custom commands are now skills, they sit alongside other Claude Code extension points such as subagents and hooks, and a skill can be invoked explicitly with its slash command or loaded automatically by Claude when relevant. Tools connected through MCP servers can also expose their own slash commands, letting you trigger external integrations the same way. For a hands-on walkthrough, see the slash commands guide.
Why it matters
Slash commands turn your repeated instructions into one-word shortcuts, so common workflows like writing tests, reviewing a diff, or scaffolding a file become consistent and instantly reusable across your team instead of being retyped from scratch each session.
Deep dive: Claude Code Slash Commands: A Practical Guide
Related terms
See Slash Commands in action across releases — browse the Claude Code changelog.
Get release alerts