Most people use Claude Code with zero idea that slash commands exist — and then wonder why it feels slow or forgetful. These little commands (you just type a /) are the difference between fighting the tool and flying. Here's a clean, accurate cheat sheet of the ones actually worth knowing, grouped by when you'd reach for them.
What's a slash command? A shortcut you type starting with / that triggers a built-in action instead of being sent to the AI as a prompt. /clear, /model, /usage — that kind of thing.
🚀 Setting up a project
| Command | What it does |
|---|---|
/init | Creates a CLAUDE.md project guide so Claude knows your stack, conventions, and key files |
/memory | Edit your CLAUDE.md files or manage auto-memory |
/add-dir <path> | Give Claude access to an additional folder during the session |
/config | Open the settings interface (alias: /settings) |
💬 Managing your session & context
| Command | What it does |
|---|---|
/clear | Start fresh with empty context (aliases: /reset, /new) |
/compact | Summarize the conversation to free up context while continuing |
/context | Visualize how much context you're using |
/resume | Resume or switch between past conversations (alias: /continue) |
/branch | Fork the conversation at the current point to explore an alternative |
/rewind | Roll back code/conversation to an earlier checkpoint (aliases: /checkpoint, /undo) |
/export | Export the conversation to a file or clipboard |
🧠 Models & usage
| Command | What it does |
|---|---|
/model | Switch the AI model and save it as your default |
/effort | Set the reasoning effort level (low / medium / high, etc.) |
/usage | Show your cost, usage limits, and stats (aliases: /cost, /stats) |
/usage-credits | Configure extra usage credits (this replaced the old /extra-usage) |
🛠️ Building & reviewing code
| Command | What it does |
|---|---|
/plan | Enter plan mode — Claude maps out the work before touching anything |
/diff | Open an interactive viewer of your uncommitted changes |
/code-review | Review your diff for bugs and cleanups (can auto-fix or comment) |
/security-review | Analyze pending changes for security vulnerabilities |
🔌 Tools, integrations & customization
| Command | What it does |
|---|---|
/mcp | Manage MCP server connections (the way Claude plugs into your other tools) |
/agents | Configure subagents |
/skills | List available skills |
/permissions | Manage allow/ask/deny rules for tools (alias: /allowed-tools) |
/theme | Change the color theme |
/help | Show help and the full list of available commands |
The 5 to learn first: /init (give it context), /clear (fresh start), /compact (trim a long chat), /model (right model for the job), and /usage (watch your limits). Those five alone fix most of the "Claude Code is slow/forgetful" complaints.
Claude Code adds and updates commands over time, and many have aliases. For the complete, current list, type /help inside Claude Code or check the official docs. This cheat sheet covers the commands most people actually use day to day.
Frequently Asked Questions
Want to actually master tools like Claude Code?
The Builders' Inner Circle takes you from "what's a slash command?" to shipping real projects — one short build a week. $4.99/mo, cancel anytime.
Join the Builders' Inner Circle →Get one free AI build in your inbox 🔨
The exact 10-minute build my paid members start with — free. No spam, unsubscribe anytime.