Iterative Refinement Loop

# Iterative Refinement Loop

You are an expert at building AI-assisted quality improvement workflows. Design a prompt loop that uses AI to generate, self-critique, and improve output across multiple cycles until it meets a defined quality standard.

## Target Output

**What I need the AI to produce:** [OUTPUT_TYPE — e.g., marketing copy, strategic plan, code function, research summary]
**Quality standard it must meet:** [QUALITY_STANDARD — e.g., a VP of Marketing would approve this without edits, the code passes all unit tests, it cites 3 credible sources per claim]
**Maximum refinement cycles:** [CYCLE_LIMIT]

## Cycle Structure

### Cycle 1: Initial Generation
Generate a first draft of [OUTPUT_TYPE] based on: [INPUT_BRIEF]

### Cycle N: Self-Critique
After each draft, critique your own work against these specific dimensions:
1. **[DIMENSION_1]:** Does this draft [CRITERION_1]? Score 1–10 and explain.
2. **[DIMENSION_2]:** Does this draft [CRITERION_2]? Score 1–10 and explain.
3. **[DIMENSION_3]:** Does this draft [CRITERION_3]? Score 1–10 and explain.

List the top [MAX_ISSUES] specific, actionable improvements needed.

### Improvement Step
Rewrite the draft, addressing every issue raised in the critique. Do not preserve flaws from the previous version. List what changed.

## Stopping Condition

Stop refining when:
- All three dimensions score ≥ [MIN_SCORE]/10, OR
- You have completed [CYCLE_LIMIT] cycles

After stopping, output the final draft, the final scores, and a one-paragraph explanation of why this version meets the quality standard.