Type / at the start of your input to see the list. Type any letters after / to filter.
CONTEXT & CONVERSATION MANAGEMENT
/clear — Wipes the conversation and starts fresh. Use this every time you switch tasks. Old context from a previous task genuinely makes me worse at the new one. (aliases: /reset, /new)
/compact [instructions] — Compresses conversation history into a summary. This is the most important command to learn. Use it proactively when context gets long, not just when I start losing track. The real power move: add focus instructions like /compact keep the database schema and error handling patterns to control what survives.
/context — Visualizes your context usage as a color grid and gives optimization suggestions. Use this to see how close you are to the limit.
/fork [name] — Creates a branch of your conversation at the current point. Useful when you want to explore two different approaches without losing your place.
/rewind — Rewind the conversation and/or your code to a previous point. If I went down the wrong path, this gets you back. (alias: /checkpoint)
/export [filename] — Exports the conversation as plain text. With a filename it writes directly to a file. Without one it gives you options to copy or save.
/copy — Copies my last response to your clipboard. If there are code blocks, it shows an interactive picker so you can grab individual blocks.
MODEL & PERFORMANCE SWITCHING
/model [model] — Switches models mid-session. Use left/right arrow keys to adjust effort level in the picker. Common pattern: start with Sonnet for routine work, flip to Opus for hard problems, switch back when you're done.
/fast [on|off] — Toggles fast mode for Opus 4.6. Faster output, same model. Good for straightforward edits.
/effort [low|medium|high|max|auto] — Sets how hard I think. This shipped quietly in a changelog and most people missed it. low and medium and high persist across sessions. max is Opus 4.6 only and session-scoped. auto resets to default.
CODE REVIEW & SECURITY
/diff — Opens an interactive diff viewer showing every change I've made. Navigate with arrow keys. Run this as a checkpoint after any series of edits — it's your chance to catch my mistakes before they compound.
/pr-comments [PR URL|number] — Shows GitHub PR comments. Auto-detects the PR or takes a URL/number.
/security-review — Analyzes pending changes for security vulnerabilities: injection, auth issues, data exposure. Run this before shipping anything sensitive.
SESSION & USAGE TRACKING
/cost — Detailed token usage and cost stats for the session (API users).
/usage — Shows plan usage limits and rate limit status.
/stats — Visualizes daily usage patterns, session history, streaks, and model preferences over time.
/resume [session] — Resume a previous conversation by ID, name, or interactive picker. (alias: /continue)
/rename [name] — Renames the session. Without a name, I auto-generate one from the conversation history.
/insights — Generates an analysis report of your Claude Code sessions — project areas, interaction patterns, friction points.
MEMORY & PROJECT CONFIG
/memory — View and edit my persistent memory files (CLAUDE.md). Enable/disable auto-memory and view auto-memory entries. If I keep forgetting something about your project, check this first.
/init — Initialize a project with a CLAUDE.md guide file. This is how you teach me about your codebase from the start.
/hooks — View hook configurations for tool events. Hooks let you run code automatically before or after I make changes.
/permissions — View or update tool permissions. (alias: /allowed-tools)
/config — Opens the settings interface for theme, model, and output style. (alias: /settings)
MCP & INTEGRATIONS
/mcp — Manage MCP server connections and OAuth authentication. MCP is how you connect me to external tools like GitHub, databases, APIs.
/ide — Manage IDE integrations (VS Code, JetBrains) and show connection status.
/install-github-app — Set up the Claude GitHub Actions app.
/install-slack-app — Install the Claude Slack app.
/chrome — Configure Claude in Chrome settings.
/plugin — Manage Claude Code plugins — install, uninstall, browse.
/reload-plugins — Reload all active plugins to apply changes without restarting.
AGENTS & TASKS
/agents — Manage subagent configurations and agent teams.
/tasks — List and manage background tasks.
/plan [description] — Enter plan mode directly from the prompt. I'll outline what I'm going to do before doing it.
/btw [question] — Ask a side question without adding it to the conversation. Works while I'm processing something else.
SESSION MANAGEMENT & CROSS-DEVICE
/desktop — Continue the session in the Claude Code Desktop app. macOS and Windows. (alias: /app)
/mobile — Show a QR code for the Claude mobile app. (aliases: /ios, /android)
/remote-control [name] — Makes the session controllable from claude.ai or the Claude app. (alias: /rc)
/add-dir [path] — Add additional working directories to the current session.
/sandbox — Toggle sandbox mode on/off.
ACCOUNT & SYSTEM
/login — Sign in to your Anthropic account.
/logout — Sign out.
/doctor — Diagnose and verify your Claude Code installation. Run this first when something breaks.
/status — Shows version, model, account, and connectivity info.
/feedback — Submit feedback to the Anthropic team. (alias: /bug)
/release-notes — View the full changelog.
/upgrade — Open the upgrade page for a higher plan tier.
/extra-usage — Configure extra usage to keep working when rate limits are hit.
/privacy-settings — View and update privacy settings (Pro/Max only).
/passes — Share a free week of Claude Code with friends (if eligible).
/stickers — Order Claude Code stickers. Yes, this is real.
DISPLAY & PERSONALIZATION
/vim — Toggle between Vim and Normal editing modes.
/color [color|default] — Set prompt bar color for the session. Options: red, blue, green, yellow, purple, orange, pink, cyan.
/theme — Change color theme including light/dark and colorblind variants.
/terminal-setup — Configure terminal keybindings for Shift+Enter. Run this if multi-line input isn't working.
/keybindings — Open or create keybindings configuration.
/statusline [description] — Configure the Claude Code statusline. Describe what you want or run it empty for auto-configuration.
/voice — Push-to-talk voice mode. Hold spacebar to speak. Supports 20+ languages.
/skills — List all available skills.
BUNDLED SKILLS (the real power moves)
These look like slash commands but are AI-driven workflows. They load specialized instructions into my context and I orchestrate multi-step processes, including spawning parallel agents:
/simplify [focus] — I review recently changed files for code reuse, quality issues, and efficiency improvements. Spawns three review agents in parallel, aggregates findings, and applies fixes automatically. Run this after every feature.
/debug [description] — Structured debugging workflow by reading the debug log. Way more effective than just saying "fix this bug."
/batch [instruction] — Orchestrates large-scale changes in parallel. I decompose the work into 5-30 units, spawn one agent per unit in an isolated git worktree, and create PRs. Example: /batch "migrate src/ from Solid to React"
/loop [interval] [prompt] — Runs a prompt repeatedly on an interval. Useful for polling deployments or monitoring PRs. Example: /loop 5m "check if deploy finished"
/claude-api — Loads Claude API and Agent SDK reference for your project language. Also activates automatically when your code imports the Anthropic SDK.