Every Claude Code Command

Picks up your last conversation where you left off. Restores context, files, and task state.

Creates a new conversation fork from the current point. Useful for exploring alternatives without losing your main thread.

Rolls back to an earlier message in the conversation. Undo mistakes without starting over.

Give your conversation a descriptive name for easier recall later.

Save the full conversation as JSON or markdown for documentation, sharing, or archival.

See how much the current session has cost in real-time. Track spending before it surprises you.

Aggregate usage stats across the session — tokens in/out, cost per model, and efficiency metrics.

Granular breakdown of token usage by tool call, message, and model. Find what is eating your budget.

Manage CLAUDE.md and memory files. View what Claude remembers about your project and edit persistent context.

Add more folders to the project scope. Useful for monorepos or when your code spans multiple directories.

Manage Claude Code settings — model preferences, permissions, allowed tools, and behavior flags.

View all changes relative to the codebase. See what Claude has modified before committing.

Dedicated security scan focused on OWASP top 10, injection vulnerabilities, auth issues, and data exposure.

Enter plan mode. Claude analyzes the task, breaks it into steps, and waits for approval before executing. Essential for 3+ step tasks.

Control which files Claude can read and write. Set boundaries for sensitive files like .env or credentials.

Reduce project data in the context window. Use when sessions get slow or context is filling up.

View all specialized agents available in your project. See their capabilities, assigned skills, and status.

Hot-reload all plugins without restarting Claude Code. Use after installing new plugins or updating configurations.

Switch between light, dark, and custom themes for the Claude Code terminal interface.

Customize the color scheme of the terminal output — text, highlights, borders, and accent colors.

Display the full list of built-in commands with descriptions. Your starting point for discovery.

Display the current version. Some features require specific versions (e.g., /ultra-review needs 2.1.86+).

View connection status, active MCP servers, loaded plugins, and current model information.

Report bugs, suggest features, or share feedback directly with the Claude Code team.

Re-read memory files, process scratchpad, update working memory, and reconcile task status.

Save learnings, update memory, archive session, clear scratchpad, and write daily note.

Rate work 1-10 on correctness, completeness, security. Spawns review agent to audit recent changes.

Analyze task, search skills, select agents, define success criteria. Enter plan mode before executing.

Review all changes on the current branch against coding standards and project conventions.

When stuck in a loop, get a fundamentally different approach. Diagnoses the root cause and suggests alternatives.

Save all state to memory files before clearing context. Resume fresh without losing progress.

Review what worked, what failed, and what to improve. Extract lessons and update the knowledge base.