Reference: Interactive mode
Type / in a session to see all available commands. Type / followed by letters to filter.
| Command | Description |
|---|---|
/init |
Generate a CLAUDE.md for the current project |
/model [model] |
Switch model (sonnet, opus, haiku, opusplan) |
/effort [level] |
Set reasoning effort: low, medium, high, max (Opus only) |
/plan |
Enter Plan Mode (analyze without modifying files) |
/clear |
Reset context window. Aliases: /reset, /new |
/compact [instructions] |
Summarize conversation to free up context |
/rewind |
Restore conversation and/or code to a previous checkpoint. Alias: /checkpoint |
/permissions |
View and edit tool permission rules. Alias: /allowed-tools |
/memory |
Edit CLAUDE.md files and manage auto-memory |
/hooks |
Manage pre/post tool-use hook configurations |
/mcp |
Manage MCP server connections |
/context |
Visualize current context usage as a colored grid |
/cost |
Show token usage statistics for the session |
/status |
Show version, model, account, and connectivity info |
/help |
List all available commands |
/exit |
Exit Claude Code. Alias: /quit |
| Command | Description |
|---|---|
/diff |
Interactive viewer for uncommitted changes and per-turn diffs |
/copy [N] |
Interactive selector to copy, paste, or write a code block from the conversation |
/review |
Review a pull request (requires gh CLI) |
/pr-comments [PR] |
Fetch comments from a GitHub PR |
/rename [name] |
Rename the current session |
/resume [session] |
Resume a past session by name or ID |
/branch [name] |
Create a branch of the current conversation. Alias: /fork |
/skills |
List available custom skills (see Custom Skills below) |
/vim |
Toggle Vim editing mode |
/theme |
Change the color theme |
/sandbox |
Toggle sandbox mode (OS-level isolation) |
/fast [on|off] |
Toggle fast mode |
/btw <question> |
Ask a quick side question without adding it to conversation history |
/security-review |
Analyze pending changes for security vulnerabilities |
/doctor |
Diagnose installation and settings issues |
/insights |
Generate a report on Claude Code sessions |
/stats |
Visualize usage and session history |
/schedule |
Create and manage cloud scheduled tasks |
/release-notes |
View the changelog |
/remote-control |
Enable remote control from claude.ai. Alias: /rc |
/reload-plugins |
Reload plugins without restarting |
Reference: Skills
Custom commands have been merged into skills. Both .claude/commands/review.md and .claude/skills/review/SKILL.md create a /review command and work the same way. Existing .claude/commands/ files keep working — no migration needed.
Skills are the recommended approach going forward, as they add more capabilities:
Old (.claude/commands/) |
New (.claude/skills/) |
|
|---|---|---|
| Format | Single .md file |
Directory with SKILL.md |
| Supporting files | No | Yes |
| Claude auto-invokes | No | Yes (via description field) |
| Invocation control | No | Yes (disable-model-invocation) |
| Subagent execution | No | Yes (context: fork) |
| Tool restrictions | No | Yes (allowed-tools) |
| Path-specific activation | No | Yes (paths) |
.claude/skills/fix-issue/SKILL.md
---
name: fix-issue
description: Fix a GitHub issue by number
disable-model-invocation: true
---
Fix GitHub issue $ARGUMENTS following our coding standards.
1. Read the issue, implement the fix, write tests, create a commit.Invoke with /fix-issue 123. The $ARGUMENTS placeholder is replaced with 123. You can also use $ARGUMENTS[N] or $N to reference individual arguments, and ${CLAUDE_SKILL_DIR} to reference the skill's directory.
| Location | Path | Scope |
|---|---|---|
| Personal | ~/.claude/skills/<name>/SKILL.md |
All your projects |
| Project | .claude/skills/<name>/SKILL.md |
This project only |