ai-prompt-tune

# Prompt Tune ## Purpose Prompt optimization and creation. Improves prompts, skill descriptions, and agent instructions using proven techniques: explicit over implicit, examples over rules, structured formatting, and positive framing. Can auto-enhance prompts for other skills or manually optimize user-provided text. ## Trigger - Command: `/ai-prompt-tune "<text>"` (optimize text) or `/ai-prompt-tune --skill <name>` (optimize a skill's description) - Context: writing a new prompt, improving an existing skill's description, crafting agent instructions. ## When to Use - Writing or refining skill `description` fields (CSO optimization) - Crafting system prompts for AI integrations - Improving agent instruction clarity - Before publishing any prompt-based artifact ## Optimization Techniques Apply these in order of impact: ### 1. Be Explicit Over Implicit | Before | After | |--------|-------| | "Handle errors properly" | "Wrap database calls in try/except, log the exception with stack trace, return a structured error response with HTTP 500" | | "Follow best practices" | "Apply guard clauses for early return, extract methods over 20 lines, name variables by intent not type" | ### 2. Show, Do Not Tell Replace rules with examples. One concrete example is worth five abstract instructions. ``` Bad: "Use descriptive names" Good: "Name variables by what they represent: - `user_count` not `n` - `is_valid` not `flag` - `retry_delay_seconds` not `delay`" ``...