happycapy-ai/happycapy-skills/prompt-improver

Prompt Improver

Transform vague prompts into clear, specific, actionable ones for better AI responses.

Workflow

1. Gather context - Use AskUserQuestion to clarify:

• Target platform (Claude Code, ChatGPT, API, image gen)
• Priority (accuracy, speed, depth, creativity)
• Missing context (technical stack, constraints, examples)

1. Analyze - Identify what's unclear, missing, or ambiguous

1. Improve - Apply the framework (see references/framework.md)

1. Present - Show improved prompt with key changes explained

1. Refine - Ask if user wants adjustments

AskUserQuestion Templates

Initial clarification:

questions:
  - header: "Platform"
    question: "What will you use this prompt for?"
    options:
      - label: "Claude Code"
        description: "Coding, file ops, terminal"
      - label: "ChatGPT/Claude.ai"
        description: "General conversation"
      - label: "API/Automation"
        description: "Programmatic use"
      - label: "Image gen"
        description: "DALL-E, Midjourney, etc."
  - header: "Priority"
    question: "What matters most?"
    options:
      - label: "Accuracy"
        description: "Correctness is critical"
      - label: "Speed"
        description: "Quick, concise"
      - label: "Depth"
        description: "Comprehensive"
      - label: "Creativity"
        description: "Novel approaches"

Post-improvement:

header: "Refine"
question: "Adjust the improved prompt?"
options:
  - label: "Looks good"
    description: "Use as-is"
  - label: "More specific"
    description: "Add constraints"
  - label: "More concise"
    description: "Shorten"
  - label: "Different focus"
    description: "Change emphasis"

Output Format

## Analysis
[Brief issues/opportunities]
 
## Improved Prompt
[Ready-to-use prompt]
 
## Key Changes
-[Change]: [Why]

Quick Mode

If user says "quick improve", skip questions and make reasonable assumptions. Note assumptions made.

Aristotelian Mode (First Principles)

Activated when user says "Aristotelian", "first principles", or "proof-based". Instead of the standard framework, produce a prompt that instructs the receiving LLM to reason from first principles when executing the task.

The prompt-improver does NOT do the Aristotelian reasoning itself. It crafts a prompt that tells the LLM to:

1. Gather context from user - Ask what system capabilities, tools, and constraints exist. Bake known context (root access, AI model, available tools, domain) directly into the prompt as given axioms.

1. Embed the reasoning directive - The improved prompt tells the LLM to:

• Identify the atomic, irreducible truths of the task before acting
• Interrogate each truth: "Can this be decomposed further? If removed, does the task break? Does it contradict anything?"
• Discard anything that is not strictly necessary
• Build the solution deductively, where every action traces to a stated axiom
• Verify the result against the axioms at the end

1. Structure the output prompt with these sections:

`` REASONING DIRECTIVE: [Instruct the LLM to use first-principles reasoning] GIVEN AXIOMS: [Known truths about system, capabilities, domain -- baked in] TASK: [What to accomplish] METHOD: [Tell LLM to discover task-specific axioms, interrogate them, then build deductively] VERIFICATION: [Tell LLM to check its result against its axioms] ``

Output format for Aristotelian mode:

## Analysis
[What context was embedded and why]
 
## Improved Prompt (Aristotelian)
[The complete prompt with reasoning directive, given axioms, task, method, and verification]
 
## What This Prompt Does
-Tells the LLM to [specific reasoning behavior]
-Bakes in [specific context] so the LLM does not hallucinate it

See references/aristotelian.md for the full methodology and prompt structure.

References

• Framework details: See references/framework.md for the 6-principle improvement framework
• Aristotelian mode: See references/aristotelian.md for the proof-based first principles methodology
• Examples: See references/examples.md for before/after transformations
• Anti-patterns: See references/anti-patterns.md for common issues to fix