Skill Store 33

Slash Commands

Reference for Claude Code slash commands, keyboard shortcuts, and input prefixes, plus the format fo

by thelobbi · MIT

The skill

Reference for Claude Code slash commands, keyboard shortcuts, and input prefixes, plus the format for writing custom command skill files (SKILL.md frontmatter, triggers, args) in .claude/skills/ or plugin commands/ directories. Use when the user asks which slash command or shortcut does something, how to create or invoke a custom command, or how command resolution and plugin namespacing work.

Claude Code Slash Commands

Complete reference for all built-in and custom slash commands.

Built-in Slash Commands

Session Management

Command Description
/clear Clear conversation and start fresh (same session)
/compact Compress conversation to reduce context usage
/compact [instructions] Compact with custom focus instructions
/resume Resume a previous session
/continue Continue current session
/status Current session info
/context View context window usage

Help & Info

Command Description
/help Show available commands
/doctor Run diagnostics and health checks
/version Show Claude Code version
/settings View active settings

Model & Configuration

Command Description
/model View or change the current model
/model <name> Switch to specific model (sonnet, opus, haiku)
/permissions Manage permission rules

Conversation Modes

Command Description
/plan Enter plan mode (read-only analysis)
/fast Toggle fast mode (faster output, same model)

Memory & Instructions

Command Description
/memory Edit CLAUDE.md and auto-memory
/init Generate project CLAUDE.md
/init-only Generate CLAUDE.md and exit

Agents & Skills

Command Description
/agents View, create, edit, delete subagents
/agents list List all available agents
/skills Browse available skills
/plugin Manage plugins (install, enable, disable)

Tools & Servers

Command Description
/mcp Check MCP server status, authenticate
/hooks Configure lifecycle hooks
/add-dir <path> Add additional working directory

Code & Navigation

Command Description
/diff View git diff interactively
/review Review pull request
/fork [name] Fork current conversation

Display & Preferences

Command Description
/theme Change color theme
/vim Toggle Vim editing mode
/terminal-setup Configure keybindings

Built-in Skills

Command Description
/simplify Review code quality and refactor
/batch <instruction> Parallel codebase changes
/debug [description] Debug session issues
/loop <interval> Recurring prompts
/claude-api Load Claude API reference docs

Security & Sandbox

Command Description
/sandbox View or configure sandbox settings
/security-review Run security analysis on codebase

Export & Sharing

Command Description
/export Export conversation
/copy Copy last response to clipboard

Monitoring & Insights

Command Description
/tasks View background tasks
/insights Session insights and analytics
/stats Usage statistics
/usage Token/cost usage for current session
/extra-usage Extended usage details

Releases & Updates

Command Description
/release-notes View release notes for current version
/upgrade Upgrade Claude Code to latest

Hooks & Plugins

Command Description
/save-hooks Save current hooks configuration
/reload-plugins Reload installed plugins

Browser & Display

Command Description
/chrome Open Chrome browser session
/desktop Desktop view mode
/mobile Mobile view mode
/ide IDE integration mode

Account & Auth

Command Description
/login Login to Anthropic
/logout Logout from Anthropic

Configuration

Command Description
/rename Rename current conversation
/keybindings View or edit keybindings
/statusline Configure status line display
/output-style Change output formatting style
/privacy-settings View or edit privacy settings
/add-dir Add working directory to context

Remote & Apps

Command Description
/remote-env Configure remote environment
/remote-control Start remote control session

GitHub & Slack Apps

Command Description
/install-github-app Install Claude GitHub App
/install-slack-app Install Claude Slack App

Bug Reporting

Command Description
/bug Report a bug (can be disabled via env var)

Input Prefixes

Prefix Action
/ Invoke command or skill
! Bash mode (execute shell command)
@ File mention (add file to context)
? Show shortcuts

Keyboard Shortcuts

Input Controls

Shortcut Action
Enter Send message
Shift+Enter Multi-line input
Ctrl+C Interrupt current operation
Ctrl+D Exit Claude Code
Escape Rewind to last checkpoint
Escape (2x) Full rewind
Tab Auto-complete
Up/Down Navigate command history
? Show shortcuts

Mode & Toggle Controls

Shortcut Action
Shift+Tab Cycle permission modes
Ctrl+O Toggle verbose output
Ctrl+B Send task to background
Ctrl+T Toggle task list
Option+P Switch model
Option+T Toggle extended thinking
Ctrl+G Open input in text editor
Ctrl+V Paste image

Line Editing

Shortcut Action
Ctrl+K Kill text from cursor to end of line
Ctrl+U Kill text from cursor to start of line
Ctrl+Y Yank (paste) killed text
Alt+Y Cycle through kill ring
Alt+B Move back one word
Alt+F Move forward one word
Alt+M Toggle multiline mode

Display & Navigation

Shortcut Action
Ctrl+L Clear screen
Ctrl+R Reverse search history
Ctrl+F Kill background agents

Permission Prompts

Shortcut Action
y / Enter Approve (Allow Once)
a Allow Always (for this session)
n Deny
d Deny Always (for this session)
e Edit (modify the tool call)

Custom Slash Commands

Creating Custom Commands

Custom commands are skill files in .claude/skills/ or plugin directories.

Skill File Format
---
name: my-command
description: What this command does
triggers:
  - /my-command
  - /mc
args:
  - name: target
    description: Target to operate on
    required: false
---

# My Custom Command

Instructions for Claude when this command is invoked.

## Steps
1. Do this first
2. Then do this
3. Finally do this

## Rules
- Always follow this pattern
- Never do this other thing
Plugin Command Format

Commands in plugin commands/ directories:

# Command Name

Description of what the command does.

## Usage
\`\`\`
/plugin-prefix:command-name [args]
\`\`\`

## Arguments
- `arg1` — Description
- `--flag` — Description

## Implementation

When this command is invoked:
1. Step one
2. Step two
3. Step three

Command Resolution Order

  1. Built-in commands (highest priority)
  2. Project skills (.claude/skills/)
  3. Installed plugin commands
  4. User-level skills (~/.claude/skills/)

Invoking Custom Commands

# Direct invocation
/my-command

# With arguments
/my-command some-argument

# Plugin-namespaced command
/plugin-name:command-name

# Via Skill tool (programmatic)
Skill(skill="my-command", args="some-argument")

Plan Mode Details

When /plan is active:

  1. Claude analyzes the request
  2. Writes a step-by-step plan to a plan file
  3. Calls ExitPlanMode to present plan to user
  4. User reviews and approves/modifies/rejects
  5. On approval, Claude executes the plan

Plan Mode Permissions

  • Read-only tools work normally
  • Write/Edit/Bash require plan approval first
  • Use AskUserQuestion for clarifications during planning
  • Use ExitPlanMode when plan is ready (not AskUserQuestion)

Context Management Commands

/compact

Compresses the conversation to free up context window space.

/compact                    # Default compression
/compact focus on the API   # Compress with focus on API-related content

When to use:

  • Context window is getting large
  • Conversation has covered many topics
  • You want to focus on a specific area

How it works:

  • Summarizes previous conversation
  • Preserves key decisions and context
  • Reduces token count significantly
  • Maintains file state awareness

/clear

Starts a completely fresh conversation.

/clear

When to use:

  • Switching to unrelated task
  • Want a clean slate
  • Context is too polluted

Difference from /compact:

  • /compact preserves summarized context
  • /clear removes all context entirely

Use this skill

Connect your agent to the Skill Store MCP server and ask for it by ref: thelobbi/slash-commands-60040518965.

Download SKILL.md · Markdown version

Details

Publisher
thelobbi
Category
AI & Agents › Skill & prompt authoring
License
MIT
Updated
2026-10-08

References

More from thelobbi