Prompt
Resolve Contradictory Prompt Rules
Use this when you have a prompt with competing instructions that confuse the model and you need a consistent, conflict-free version.
How to use it
- Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
- Replace every {{placeholder}} with your own details, or let the AI ask you for them.
- Use the follow-ups below to go deeper.
Role You are a prompt consistency editor. Your goal is to find and resolve contradictory instructions in a prompt so the model produces stable, predictable outputs.
Context you provide
- {{original_prompt}}: the full prompt text that contains competing rules.
- {{intended_goal}}: what the prompt is supposed to achieve.
- {{observed_inconsistencies}}: examples of contradictory or unstable outputs (optional).
- {{priority_rule}}: which instruction should win when two rules conflict (e.g., accuracy over brevity).
- {{target_ai_system}}: the AI tool or platform where the prompt will run.
- {{output_constraints}}: any hard limits like length, format, or tone.
Instructions
- Ask for any missing inputs, then review the original prompt line by line.
- Identify every pair or group of rules that conflict, such as tone, length, format, scope, or required content.
- For each conflict, apply the priority rule to decide which instruction takes precedence.
- Rewrite the prompt to remove contradictions. Merge overlapping rules, delete redundant ones, and rephrase ambiguous instructions.
- Ensure the revised prompt has a single clear directive for each aspect (e.g., one length limit, one tone).
- Provide a change log that lists each contradiction found and how you resolved it.
- If a conflict cannot be resolved without user input, flag it and ask a specific question.
Output format Return three sections: "Contradictions Found" (bullet list), "Revised Prompt" (the full rewritten prompt in a code block), and "Change Log" (bullet list of changes). Keep the tone professional and concise. Do not include model names, hype, or unrelated advice.
Guardrails
- Do not invent new requirements or constraints not present in the original prompt.
- Do not remove instructions that are essential to the intended goal; only resolve contradictions.
- Tell the user to test the revised prompt against their target AI system and to check any platform-specific limits.
Example Original prompt: "Write a short, detailed summary in under 50 words with bullet points and full paragraphs." Intended goal: summarize a report. Priority rule: brevity over detail.