⚡ EXECUTION INSTRUCTIONS FOR CLAUDE

When this command is invoked, YOU (Claude) must execute these steps immediately: This is NOT documentation - these are COMMANDS to execute right now. Use TodoWrite to track progress through multi-phase workflows.

🚨 EXECUTION WORKFLOW

Phase 1: Execute Documented Workflow

Action Steps:

1. Review the reference documentation below and execute the detailed steps sequentially.

📋 REFERENCE DOCUMENTATION

Combinations Command

Usage: /combinations or /combos

Description: Lists all available slash command combinations and their effects.

Available Command Combinations

True Universal Command Composition

BREAKTHROUGH: The system now leverages Claude's natural language processing for genuine universality!

How It Works

Instead of complex hardcoded mappings, the system creates simple meta-prompts that Claude interprets naturally:

Input: /think /debug /weird analyze performance Meta-prompt: Use these approaches in combination: /think /debug /weird. Apply this to: analyze performance Claude's Natural Understanding: Combines deep thinking, systematic debugging, and unconventional approaches

Revolutionary Simplicity

• ANY commands work - even completely made-up ones
• No predefined lists - Claude understands context and meaning
• Consistent quality - no degradation for "unknown" commands
• Self-improving - gets better as Claude's understanding evolves

Examples of True Universality

Real Commands:

• /think /plan /arch → Strategic architectural planning with deep thinking
• /debug /optimize /test → Systematic debugging with optimization and testing

Creative Commands:

• /mythical /dragon /optimize → Creative powerful optimization approaches
• /stealth /ninja /implement → Subtle, efficient implementation strategies

Completely Made-Up Commands:

• /quantum /cosmic /analyze → Claude interprets creatively for analysis
• /fluffy /rainbow /debug → Claude finds meaningful interpretation

Technical Commands:

• /security /scale /deploy → Security-focused scalable deployment
• /performance /monitor /validate → Performance monitoring with validation

How It Works

The Universal Command Composition system:

1. Parses ANY commands from your input (known or unknown)
2. Builds natural language instructions dynamically
3. Combines behaviors intelligently without predefined rules
4. Handles unknown commands gracefully with "approach" patterns

Dynamic Examples

Input: /think /debug /brief /test analyze memory leaks Output: /execute Use sequential thinking for systematic debugging with testing methodology: analyze memory leaks [Respond concisely]

Input: /arch /optimize /secure build API gateway Output: /arch for architectural design to optimize and improve with security considerations: build API gateway

Input: /custom /weird /nonexistent create something Output: /execute with custom approach with weird approach with nonexistent approach: create something

Input: /research /validate /monitor /verbose setup CI/CD Output: /execute with research methodology to validate and verify to monitor and track: setup CI/CD [Provide detailed explanations]

Configuration

The system requires no configuration! Claude's natural language processing handles all command interpretation. The hook script at .claude/hooks/compose-commands.sh is simple and needs no customization.

Usage Notes

• Order doesn't matter: /think /plan = /plan /think
• Single commands work unchanged: /think remains /think
• Unknown combinations pass through unchanged
• Conflicting combinations (like /brief /verbose) are handled gracefully