Claude Code is driven from the prompt with slash commands. This guide groups 35 of the most useful ones by purpose, so you can find the right command quickly.
Commands evolve between releases. Type / in a session to see the list available in your version, and check the official commands reference for arguments and aliases.
Control the session
| Command | What it does |
/context | Visualize what is using your context window, with optimization suggestions |
/clear | Start a new conversation with empty context |
/compact [instructions] | Summarize the conversation to free context while keeping the essentials |
/resume [session] | Reopen a previous session by ID or name, or pick one from a list |
/branch [name] | Branch the conversation at this point to try another direction without losing the original |
/rewind | Roll the conversation and/or code back to a previous point (aliases: /checkpoint, /undo) |
/fork [prompt] | Copy the conversation into a new background session and keep working here |
| Command | What it does |
/init | Generate a starter CLAUDE.md for the project |
/memory | Edit CLAUDE.md files and manage auto memory |
/cd | Move the session to another working directory, keeping the conversation |
/permissions | Manage allow, ask and deny rules for tools |
/add-dir | Give Claude access to an additional directory for this session |
/terminal-setup | Install a Shift+Enter newline keybinding in supported terminals and editors |
/mcp | Manage MCP server connections and authentication |
Utilities
| Command | What it does |
/usage | Show session cost, plan limits and activity stats |
/status | Open the Status tab: version, model, account and connectivity |
/debug [description] | Enable debug logging and troubleshoot from the session log |
/doctor | Run a setup checkup on installation, settings and CLAUDE.md, and propose fixes |
/feedback [report] | Send product feedback or report a bug |
Tune Claude
| Command | What it does |
/voice | Dictate prompts instead of typing (requires a Claude.ai account) |
/model [model] | Switch the model |
/effort [level] | Set how much reasoning effort Claude spends |
/color [color] | Tint the prompt bar for the current session |
/config | Open settings: theme, model, output style and more |
/keybindings | Open your keyboard shortcuts file |
Code better
| Command | What it does |
/plan [description] | Enter plan mode to design the work before any edit |
/diff | Review the changes in your working tree, including Claude’s edits |
/code-review | Review the current diff, a PR, a branch or a path for correctness bugs |
/security-review | Analyze your branch changes for vulnerabilities such as injection and auth issues |
Work in parallel
| Command | What it does |
/tasks | View and manage background work, including finished subagents |
/btw [question] | Ask a side question without adding it to the conversation |
/batch | Split a large change into independent units handled by parallel background agents |
/background [prompt] | Detach the session to run as a background agent and free the terminal |
/agents | Create or manage subagents (stored in .claude/agents/) |
/remote-control | Make the session available from claude.ai to follow it remotely (alias: /rc) |
A suggested workflow
- Run
/init once, then refine the generated CLAUDE.md (see Project Structure). - Start a task with
/plan so Claude proposes an approach before editing. - Watch
/context during long tasks and run /compact before the window fills. Use /clear between unrelated tasks. - Use
/rewind or /branch when an approach goes wrong or you want to compare alternatives. - Before committing, run
/diff, /code-review and /security-review. - Use
/permissions to pre-approve safe commands so Claude asks less often.
For a first-time setup walkthrough, see Setting Up a New Project with Claude Code.
