Commands Cheat Sheet
Comprehensive list of all available commands
/command and let AI do the heavy lifting! No coding required, just tell ClaudeKit what you want. Example
/cook "add a login button" → AI builds the entire feature for you! 🎉 Your Coding Journey in 3 Simple Steps
From idea to working code — no programming knowledge needed!
Start with an Idea
Have an idea? Just talk to ClaudeKit like you would to a friend!
/brainstorm "I want to add a dark mode toggle to my website" Or natural language: "brainstorm ideas for [feature]"
Create a Plan
AI creates a detailed roadmap for building your feature
/plan @brainstorm_report.md - What files to create/modify
- Step-by-step instructions
- How to test everything works
/clear after reviewing your plan and before implementing to free up context. This is now mandatory in v2.5.0+. AI Builds It!
Sit back and watch AI write the code, test it, and make it perfect!
/cook @plan.md Auto-detects plan path → code mode
Or: "implement @plan.md"
/code @plan.md Implement from existing plan
Use /cook in engineer@v2.5.0+
- Writes all the code
- Runs tests to verify it works
- Reviews code quality
engineer@v2.5.0+ Breaking ChangesCommands → Skills: Many commands are now skills. Invoke via /skill-name (Claude Code v2.1.1+) or natural language.
Mandatory: Run /clear after /plan before implementing to free context.
Migration Reference
| Use Instead | Example | Previous |
|---|---|---|
/git skill | /git commit | /git:cm, /git:cp, /git:merge, /git:pr |
/brainstorm skill | /brainstorm [feature] | /brainstorm cmd, brainstorming skill |
/scout skill | /scout [directory] | scout agent removed |
/cook skill | /cook [task] | /cook:*, /code:* |
/fix skill | /fix [issue] | /fix:* cmd, fixing skill |
/remotion skill | /remotion [description] | /design:video renamed |
/threejs skill | /threejs [scene] | /design:3d renamed |
/frontend-design skill | /frontend-design [component] | /design:describe, :fast, :good, :screenshot |
/copywriting skill | /copywriting [page] | /content:* |
/skill-creator skill | /skill-creator [topic] | /skill:* |
/payment-integration skill | /payment-integration stripe or polar, sepay | /integrate:* |
Removed Agents (v2.6.0)
These agents were removed in engineer@v2.6.0. Use Task tool with appropriate subagent_type or skills instead.
copywriter database-admin scout (skill: /scout ✓) scout-external /cook --auto v2.7.x The /cook skill changed its default approval behavior between versions.
v2.6.x: auto v2.7.x: manual (add --auto if you want auto-approval) /cook --interactive v2.9.0+ --interactive is now the default mode in v2.9.0+. Full workflow with user input at each step. Uses native Claude Tasks API for task management.
v2.9.0: --interactive (default) Uses native Claude Tasks (TaskCreate, TaskUpdate, TaskGet, TaskList) /find-skills v2.9.0+ skill Use when user asks "how do I do X", "find a skill for X", or wants to extend agent capabilities. Skills CLI is the package manager for open agent skills ecosystem.
npx skills find [query] npx skills add <package> npx skills check npx skills update Browse skills: skills.sh
/plan → Validate v2.9.0+ Post-Plan gate now includes a "Validate" option to run /plan:validate before implementation. Interview-style confirmation of plan decisions.
Claude Code v2.1.1+: Skills invocable via /skill-name [args] or natural language
Token Consumption /skill âš¡
âš¡ = Token Consumption - Estimated context size when skill is activated. For reference only.
New Skills
New or converted from commands
/cook skill Task orchestration
Intent detection, subagent patterns, review cycles
--interactive--fast--parallel--no-test--auto /scout skill Codebase file discovery
Internal and external scouting
ext /fix skill Intelligent fix routing
Auto-routes based on issue type
/web-testing skill E2E, visual, a11y, perf testing
Playwright, WCAG, Core Web Vitals
/git skill Unified git operations
commit, push, PR, merge
/remotion skill React-based video creation
Animations, 3D graphics, captions
/payment-integration skill Stripe, Polar & SePay payments
Subscriptions, checkout, VietQR, webhooks
/gkg skill Semantic code analysis
GitLab Knowledge Graph
/brainstorm skill Solution brainstormer
Brutal honesty mode
/chrome-devtools skill Browser automation with Puppeteer
AI-optimized browser automation for screenshots, testing, and web scraping
/agent-browser skill AI browser agent automation
AI-optimized browser agent for complex web interactions
/shader skill GLSL fragment shaders
Write GLSL fragment shaders. Works with Three.js, WebGL, and 3D graphics projects. Use with /threejs for full 3D workflow.
/find-skills skill Discover and install skills
Search open agent skills ecosystem, install packages, check updates
findaddcheckupdate Enhanced Skills
Significantly improved in engineer@v2.5.0+
/frontend-design skill Polished frontend interfaces
3D, immersive, quick workflows
/threejs skill 3D with Three.js
550+ examples, shaders, materials
/ai-artist skill AI prompt crafting (Nano Banana)
3500+ prompts database, styles, techniques
--mode--skip--dry-run--aspect-ratio /skill-creator skill Create/update skills
Quality criteria validation
/devops skill K8s documentation
kubectl, Helm, security
Platform Features
New engineer@v2.5.0+ platform capabilities
Native Tasks API Claude Code Tasks API
Native task management
Chrome DevTools Auth Browser auth & cookies
Persistent sessions, cookie injection
OpenCode Compat OpenCode compatible
Plural folder names
Hooks Config Toggle Enable/disable hooks
Toggle hooks via config file
Statusline Modes Display modes
Configurable statusline display
Python Venv Support Python venv support
Hooks now allow venv creation - no longer blocked
All Commands by Category
Getting Started
Bootstrap new projects quickly
/bootstrap Start a new project step-by-step with guidance
You want to create a project and prefer interactive guidance
/bootstrap:auto Start a new project automatically
You trust Claude to set up a standard project for you
/bootstrap:auto:fast Quick setup, no research
You need a basic project structure fast
/bootstrap:auto:parallel Setup with parallel execution (faster, more tokens)
Large project, you want speed and have token quota
Planning & Research
Plan, research and brainstorm
/plan Create an implementation plan with smart enhancement
Complex feature, need to figure out how to build it
/plan:fast Quick plan, no research (analyze existing code only)
Simple feature, you know the codebase well
/plan:hard Deep research + detailed plan
Complex feature, new tech, or unfamiliar codebase
/plan:two Get 2 different approaches with trade-offs
You want options before deciding
/plan:parallel Plan with parallel-executable phases
Large feature that can be split into independent parts
/plan:validate Interview you to confirm plan decisions
Before coding, make sure assumptions are correct
/plan:ci Fix GitHub Actions failures
CI/CD is broken, need a fix plan
/plan:cro Create conversion rate optimization plan
Want to improve content for better conversions
/plan:archive Write journal + archive old plans
Project done, want to document learnings
/ask Get technical/architectural answers
"How does Z work?" or "Best practice for W?"
Codebase file discovery
Internal and external scouting
ext Solution brainstormer
Brutal honesty mode
Implementation
Write codes and implementation
Task orchestration
Intent detection, subagent patterns, review cycles
--interactive--fast--parallel--no-test--auto Debugging & Testing
Fix bugs and run tests
Intelligent fix routing
Auto-routes based on issue type
E2E, visual, a11y, perf testing
Playwright, WCAG, Core Web Vitals
/debug Debug technical issues (deeper analysis)
Stuck on a bug, need detailed investigation
/test Run tests and analyze results
Want to check if everything works
/test:ui Run UI tests on website
Need to test visual/user interactions
Git Workflow
Version control and git operations
Unified git operations
commit, push, PR, merge
cmcpprmerge /worktree Git worktree management
Superproject detection, safety limits
Documentation
Manage project documentation
/docs:init Analyze codebase + create initial docs
New project or no docs exist yet
/docs:update Analyze codebase + update existing docs
Code changed, docs need to catch up
/docs:summarize Analyze + summarize in docs
Want high-level overview documentation
Design
Create beautiful designs and interfaces
Polished frontend interfaces
3D, immersive, quick workflows
React-based video creation
Animations, 3D graphics, captions
3D with Three.js
550+ examples, shaders, materials
Content Creation
Content generation and copywriting
Conversion copywriting
Headlines, email, landing pages
AI prompt crafting (Nano Banana)
3500+ prompts database, styles, techniques
Integration
Connect with 3rd party services
Stripe, Polar & SePay payments
Subscriptions, checkout, VietQR, webhooks
Authentication
OAuth, 2FA, passkeys, sessions
Code Quality
Review and improve code quality
/review:codebase Scan and analyze entire codebase
Want code health check or architectural review
/review:codebase:parallel Deep analysis with parallel verification
Large codebase, thorough review needed
Semantic code analysis
GitLab Knowledge Graph
Skills & Tools
Extend Claude's capabilities
Create/update skills
Quality criteria validation
Discover and install skills
Search open agent skills ecosystem, install packages, check updates
findaddcheckupdate /ccs Delegate task with intelligent agent selection
Want Claude to pick the right agent for the job
Utilities
Miscellaneous tools and helpers
/ck-help Show this help guide
Need command reference or forgot something
/journal Write technical journal entries
Document failures, lessons learned, project wrap-up
/kanban Open AI orchestration dashboard
Visualize plan progress, agent activity
/preview View markdown/files in beautiful reader
Want to read plans/docs in nice UI
/watzup Review recent changes + wrap up
End of session, see what was done
/use-mcp Use tools from MCP servers
Have MCP servers configured, want to use their tools
/coding-level Set coding experience level (0-5)
0=ELI5, 1=Junior (0-2y), 2=Mid (3-5y), 3=Senior (5-8y), 4=Lead (8-10y), 5=God (default)
Command Modifiers Explained
| Modifier | Meaning | Token Impact | Example |
|---|---|---|---|
| :fast | Skip research, quick execution | Lower | /plan:fast, /bootstrap:auto:fast |
| :hard | Deep analysis, thorough | Higher | /plan:hard |
| :parallel | Use multiple agents (faster) | Higher | /plan:parallel, /bootstrap:auto:parallel |
| :two | Get 2 options/approaches | Higher | /plan:two |