Prompt Engineering · Techniques

Guardrail / safety wrapper - Draft, critique, improve

Prompt Engineering - Techniques. Compatible with all major AI assistants - ChatGPT, Claude, Gemini, Grok, DeepSeek, Perplexity and Copilot.

Write a wrapper prompt that keeps responses on-topic, declines unsafe requests, and never reveals the system prompt.

Do this in three passes in a single reply:
1. DRAFT: produce a strong first version.
2. CRITIQUE: score it against 5 quality criteria and name the weak points.
3. FINAL: deliver an improved version that fixes them.
Compatible with all major AI assistants - ChatGPT, Claude, Gemini, Grok, DeepSeek, Perplexity and Copilot.

My details: [describe your specifics here]

📋 0 copies · works with ChatGPT,Claude,Gemini,Grok,DeepSeek,Perplexity,Copilot

#prompt engineering #techniques #expert

🎓 Learn the craft behind this prompt

The Prompt Engineering Academy teaches why prompts like this work — 40-lesson certified programs, free.

Browse Academy Programs

More Prompt Engineering prompts Meta-prompt to refine any prompt Meta-prompt to refine any prompt - Step-by-step Meta-prompt to refine any prompt - With examples Meta-prompt to refine any prompt - Expert framework Meta-prompt to refine any prompt - Quick version Meta-prompt to refine any prompt - Draft, critique, improve