| Command | Description | Usage Examples |
|---|---|---|
claude |
Start interactive REPL session |
claudeclaude "explain this project"
|
claude -p "query" |
Query via SDK then exit (headless mode) |
claude -p "fix the bug in auth.js"claude -p "analyze performance issues"
|
cat file | claude -p |
Process piped content |
cat error.log | claude -p "explain these errors"tail -f app.log | claude -p "monitor for issues"
|
claude -c |
Continue most recent conversation |
claude -cclaude -c -p "add more tests"
|
claude -r "id" "query" |
Resume specific session by ID |
claude -r "abc123" "finish the PR"claude -r "session-42" "continue refactoring"
|
claude update |
Update to latest version | claude update |
claude mcp |
Configure Model Context Protocol servers |
claude mcp add github npx @modelcontextprotocol/server-githubclaude mcp listclaude mcp remove github
|
| Flag | Description | Usage Examples |
|---|---|---|
--add-dir |
Add additional working directories |
claude --add-dir ../lib ../sharedclaude --add-dir /path/to/other/project
|
--allowedTools |
Whitelist tools without prompting |
claude --allowedTools "Read" "Write" "Bash(git *)"claude --allowedTools "Edit(*.py)"
|
--disallowedTools |
Blacklist specific tools |
claude --disallowedTools "Bash(rm *)" "Write(.env)"claude --disallowedTools "Delete"
|
--print, -p |
Print response and exit |
claude -p "explain this function"claude -p "generate unit tests"
|
--output-format |
Set output format (text/json/stream-json) |
claude -p "analyze code" --output-format jsonclaude -p "get metrics" --output-format stream-json
|
--verbose |
Enable verbose logging |
claude --verboseclaude -p "debug issue" --verbose
|
--max-turns |
Limit conversation turns |
claude -p "refactor" --max-turns 3claude --max-turns 5
|
--model |
Specify AI model |
claude --model sonnetclaude --model claude-3-5-opus-20241022
|
--dangerously-skip-permissions |
Skip all permission prompts (use cautiously) | claude --dangerously-skip-permissions (only in containers) |
| Command | Description | Usage Examples |
|---|---|---|
/bug |
Report bugs to Anthropic | /bug |
/clear |
Clear conversation history | /clear |
/compact [instructions] |
Compress conversation with optional focus |
/compact/compact focus on auth changes/compact keep recent edits
|
/config |
View/modify configuration |
/config/config model/config permissions
|
/cost |
Show token usage statistics | /cost |
/doctor |
Check Claude Code installation health | /doctor |
/help |
Get usage help and list commands |
/help/help permissions
|
/init |
Initialize project with CLAUDE.md | /init |
/login |
Switch Anthropic accounts | /login |
/logout |
Sign out from Anthropic account | /logout |
/memory |
Edit CLAUDE.md memory files | /memory |
/model |
Select or change AI model |
/model/model sonnet/model opus
|
/permissions |
View or update tool permissions |
/permissions/permissions allow Write/permissions deny Bash(rm *)
|
/review |
Review code or pull requests |
/review/review PR #123
|
/status |
Show current session status | /status |
/vim |
Enable vim-style editing | /vim |
/add-dir |
Add directories mid-session |
/add-dir ../backend/add-dir /shared/libs
|
/ui / /21 |
Generate UI components using 21st.dev patterns |
/ui login form/21 responsive navbar/ui data table --accessible
|
/logo |
Search and insert company logos |
/logo GitHub/logo Discord --format TSX
|
| MCP Server | Purpose | Usage Examples |
|---|---|---|
Context7 |
Official library documentation and framework patterns |
claude --c7 "React useEffect patterns"claude --context7 "Next.js API routes"
|
Magic |
Modern UI component generation from 21st.dev |
claude --magic "create responsive dashboard"claude --magic "accessible form components"
|
Sequential |
Complex multi-step reasoning and analysis |
claude --seq "debug authentication flow"claude --sequential "analyze system architecture"
|
Morphllm |
Pattern-based code editing and transformations |
claude --morph "update all React classes to hooks"claude --morphllm "enforce ESLint rules"
|
Playwright |
Browser automation and E2E testing |
claude --play "test login flow"claude --playwright "accessibility validation"
|
Serena |
Semantic code understanding and project memory |
claude --serena "rename function across project"/sc:load --deep
|
Chrome-DevTools |
Performance analysis, debugging, and browser inspection |
claude --chrome-devtools "analyze page performance"claude --devtools "debug layout shift issues"
|
Tavily |
Web search and real-time research with multi-source analysis |
claude --tavily "latest TypeScript features 2024"/sc:research "AI development trends"
|
| Flag | Description | Usage Examples |
|---|---|---|
--think |
Enable structured analysis (~4K tokens) |
claude --think "analyze architecture""debug this issue" --think
|
--think-hard |
Deep analysis with Sequential MCP (~10K tokens) |
claude --think-hard "system redesign""performance optimization" --think-hard
|
--ultrathink |
Maximum depth analysis (~32K tokens) |
claude --ultrathink "legacy modernization""critical system debugging" --ultrathink
|
--uc / --ultracompressed |
Symbol-enhanced communication (30-50% token reduction) |
claude --uc "large codebase analysis"claude --ultracompressed --scope project
|
--delegate |
Enable sub-agent parallel processing |
claude --delegate auto "refactor entire project"claude --delegate files "analyze 50+ files"
|
--concurrency [n] |
Control max concurrent operations (1-15) |
claude --concurrency 5 "parallel analysis"claude --concurrency 10 --delegate
|
--all-mcp |
Enable all MCP servers for maximum capability |
claude --all-mcp "complex multi-domain problem"claude --all-mcp --ultrathink
|
--no-mcp |
Disable all MCP servers, native tools only |
claude --no-mcp "simple task"claude --no-mcp --performance-priority
|
| Command | Description | Usage Examples |
|---|---|---|
/sc:implement |
Feature implementation with AI guidance |
/sc:implement user authentication/sc:implement --type component LoginForm/sc:implement payment --with-tests --iterative
|
/sc:build |
Build and compilation |
/sc:build/sc:build --type prod --optimize/sc:build src/ --clean
|
/sc:design |
Design and planning |
/sc:design microservices-migration/sc:design dashboard --persona-architect
|
/sc:analyze |
Multi-dimensional code analysis |
/sc:analyze --focus security/sc:analyze src/ --focus performance/sc:analyze --scope module auth/
|
/sc:troubleshoot |
Problem investigation |
/sc:troubleshoot "login not working"/sc:troubleshoot --logs error.log/sc:troubleshoot performance --focus database
|
/sc:explain |
Educational explanations |
/sc:explain React hooks/sc:explain --persona-mentor async patterns/sc:explain --answer-only JWT tokens
|
/sc:improve |
Code improvement and refactoring |
/sc:improve messy-file.js/sc:improve --safe-mode legacy//sc:improve --focus accessibility components/
|
/sc:test |
Testing operations |
/sc:test/sc:test --coverage/sc:test --type unit --magic --play
|
/sc:cleanup |
Code cleanup and maintenance |
/sc:cleanup/sc:cleanup --type imports/sc:cleanup legacy/ --preview --validate
|
/sc:document |
Documentation generation |
/sc:document README --type guide/sc:document api/ --format markdown/sc:document --persona-scribe
|
/sc:git |
Git operations |
/sc:git commit --smart-commit/sc:git status/sc:git --interactive
|
/sc:estimate |
Project estimation |
/sc:estimate "redesign auth system"/sc:estimate migration --persona-architect/sc:estimate --type complexity
|
/sc:task |
Task management |
/sc:task create "implement feature"/sc:task list/sc:task complete task-123
|
/sc:index |
Project indexing |
/sc:index/sc:index --deep/sc:index src/
|
/sc:load |
Context loading |
/sc:load --deep --summary/sc:load project-overview/sc:load --focus architecture
|
/sc:spawn |
Sub-agent creation |
/sc:spawn security-expert/sc:spawn --type analyzer code-review
|
/sc:business-panel |
Multi-expert business analysis with 9 business thought leaders |
/sc:business-panel @strategy_doc.pdf/sc:business-panel "market analysis" --experts porter,christensen/sc:business-panel --mode debate --focus innovation
|
/sc:save |
Save session context and project memory |
/sc:save session_name/sc:save --checkpoint --summary
|
/sc:brainstorm |
Interactive requirements discovery and ideation |
/sc:brainstorm "AI project management tool"/sc:brainstorm --strategy systematic --depth deep/sc:brainstorm --strategy agile --parallel
|
/sc:reflect |
Task reflection and validation analysis |
/sc:reflect --type task --analyze/sc:reflect --type session --validate/sc:reflect --type completion
|
/sc:workflow |
Implementation workflow generation |
/sc:workflow "authentication system" --strategy agile/sc:workflow @prd.md --strategy systematic --depth deep/sc:workflow enterprise-project --validate
|
/sc:spec-panel |
Expert specification review with multiple specialists |
/sc:spec-panel @api.spec.yml --mode critique/sc:spec-panel "user story" --mode discussion/sc:spec-panel @system.spec --focus architecture
|
/sc:select-tool |
Intelligent MCP tool selection and optimization |
/sc:select-tool "rename across 10 files" --analyze/sc:select-tool "bulk code updates" --explain/sc:select-tool "save project context"
|
/sc:research |
Deep research with systematic investigation and evidence-based reasoning |
/sc:research "latest React best practices"/sc:research "security vulnerabilities in Node.js" --comprehensive/sc:research "AI model comparison" --sources academic
|
| Expert | Expertise | Key Frameworks | Usage Context |
|---|---|---|---|
| 🔨 Christensen | Innovation & Disruption | Jobs-to-be-Done, Disruption Theory | Product innovation, market entry strategy |
| 📊 Porter | Competitive Strategy | Five Forces, Value Chain Analysis | Industry analysis, competitive positioning |
| 🧭 Drucker | Management Principles | Management by Objectives, Innovation | Organizational effectiveness, strategic planning |
| 🎪 Godin | Marketing & Tribes | Purple Cow, Permission Marketing | Brand differentiation, customer engagement |
| 🌊 Kim & Mauborgne | Blue Ocean Strategy | Value Innovation, Four Actions Framework | Market creation, differentiation strategy |
| 🚀 Collins | Organizational Excellence | Good to Great, Built to Last | Performance transformation, culture building |
| 🛡️ Taleb | Risk & Uncertainty | Antifragility, Black Swan Theory | Risk management, resilience planning |
| 🕸️ Meadows | Systems Thinking | Leverage Points, System Dynamics | Complex problem solving, organizational change |
| 💬 Doumont | Communication Excellence | Cognitive Load Theory, Message Clarity | Presentation optimization, documentation clarity |
| Persona | Trigger Flag | Auto-Activation Context | Specialization |
|---|---|---|---|
| 🏗️ architect | --persona-architect |
System design, architecture tasks | Long-term architecture, scalability, system patterns |
| 🎨 frontend | --persona-frontend |
UI/component work | React/Vue/Angular, accessibility, UI/UX |
| ⚙️ backend | --persona-backend |
API/server tasks | APIs, databases, server logic, microservices |
| 🛡️ security | --persona-security |
Security analysis | OWASP, vulnerabilities, threat modeling |
| 🔍 analyzer | --persona-analyzer |
Debugging, investigation | Root cause analysis, pattern detection |
| 👨🏫 mentor | --persona-mentor |
Educational tasks | Teaching, explanations, best practices |
| 🔄 refactorer | --persona-refactorer |
Code improvement | Refactoring patterns, code smells |
| ⚡ performance | --persona-performance |
Optimization tasks | Profiling, caching, query optimization |
| ✍️ scribe | --persona-scribe |
Documentation | Technical writing, documentation patterns |
| Pattern | Description | Usage Examples |
|---|---|---|
/sc:load --deep |
Load comprehensive project context |
/sc:load --deep --summary/sc:load project-overview/sc:load --focus architecture
|
/sc:save checkpoints |
Create session restoration points |
/sc:save --checkpoint "auth-complete"/sc:save session_name --summary
|
Cross-session patterns |
Maintain context between sessions |
/sc:load → Work → /sc:save → Resume laterSession lifecycle management
|
| Command | Description | Usage Examples |
|---|---|---|
superclaude --verify |
Check system dependencies |
superclaude --verifysuperclaude --verify --verbose
|
superclaude commit |
Generate AI commit messages |
superclaude commitsuperclaude commit --interactivesuperclaude commit "Add OAuth"
|
superclaude changelog |
Create smart changelogs |
superclaude changelogsuperclaude changelog --verbose
|
superclaude readme |
Auto-generate README | superclaude readme |
superclaude review |
AI code review analysis |
superclaude reviewsuperclaude review --verbose
|
superclaude docs |
Generate technical documentation | superclaude docs |
superclaude brainstorm |
Feature ideation and suggestions | superclaude brainstorm |
superclaude annotate |
Add AI notes to git history |
superclaude annotatesuperclaude annotate --verbose
|
# Start Claude Code claude # Load project context /init # Explore codebase "Show me how authentication works" # Plan changes "Think hard about refactoring the auth module" # Implement changes "Refactor authentication to use JWT tokens" # Test changes /sc:test --coverage # Commit with AI message superclaude commit --interactive
# Identify issue /sc:troubleshoot "users can't login" --logs error.log # Analyze affected code /sc:analyze auth/ --focus security # Fix with validation /sc:improve auth.js --safe-mode --preview # Verify fix /sc:test auth/ --coverage # Document changes /sc:document --type changelog
# Design phase /sc:design "payment processing system" --persona-architect # Estimate effort /sc:estimate "payment feature" --type complexity # Implementation /sc:implement payment --type service --with-tests --iterative # Security review /sc:analyze payment/ --focus security --persona-security # Performance optimization /sc:improve payment/ --focus performance # Final documentation /sc:document payment/ --format markdown
# Load PR context claude -p "Review PR #123" # Deep analysis /sc:analyze --focus quality --scope project # Security check /sc:analyze --focus security --persona-security # Performance review /sc:analyze --focus performance # Generate review report superclaude review --verbose
# Terminal 1: Frontend claude --add-dir frontend/ /sc:implement dashboard --framework react # Terminal 2: Backend claude --add-dir backend/ /sc:implement api --type api # Terminal 3: Testing claude --add-dir tests/ /sc:test --type integration
{
"model": "claude-3-5-sonnet-20241022",
"maxTokens": 4096,
"permissions": {
"allowedTools": ["Read", "Write", "Bash(git *)", "Edit"],
"deny": ["Read(.env)", "Bash(rm *)", "Write(production.*)"]
},
"hooks": {
"PostToolUse": [{
"matcher": "Write(*.py)",
"hooks": [{
"type": "command",
"command": "python -m black $file"
}]
}]
}
}
## Project Context - Stack: React, TypeScript, PostgreSQL - Architecture: Microservices ## Commands - Build: npm run build - Test: npm test - Deploy: npm run deploy ## Standards - Use TypeScript for all new code - Minimum 80% test coverage - Follow ESLint configuration
{
"mcpServers": {
"github": {
"command": "npx",
"args": ["-y", "@modelcontextprotocol/server-github"],
"env": {
"GITHUB_TOKEN": "ghp_your_token"
}
}
}
}
/clear regularly to manage context/compact for long sessions@--dangerously-skip-permissions outside containers/permissions regularly/cost command/compact to reduce token usage--uc flag for large projects.claude/commands/ in version control