After testing thousands of prompts across ChatGPT, Claude, and other AI models, I've distilled what works into a simple 5-part framework. This isn't theory — it's what I use for every prompt I write.
The Framework
Every effective prompt needs these five elements:
- Role — Who should the AI be?
- Task — What exactly do you want?
- Context — What background info matters?
- Constraints — What should it avoid?
- Format — How should the output look?
Why This Works
Most prompts fail because they're too vague. The AI doesn't know who it's supposed to be, what format to use, or what boundaries to respect. This framework forces you to be specific.
A Real Example
Here's a prompt that uses all five elements:
Role: You are an expert crypto analyst with 10 years of trading experience.
Task: Analyze [COIN] and provide a trading recommendation.
Context: The user has a $10,000 portfolio and prefers moderate-risk investments. They've been holding for 3 months.
Constraints: Do not provide financial advice. Focus on technical analysis and market sentiment only.
Format:
- Current price and 24h change
- Key support/resistance levels
- RSI and MACD indicators
- Risk assessment (1-10)
- One-paragraph summaryThe Key Insight
The biggest mistake people make is asking for something without telling the AI how to deliver it. The Format section is often what separates a useful response from gibberish.
Try It Now
Take your next prompt and run it through this framework. I guarantee you'll see better results — and if you don't, the structured format makes it easy to identify what needs adjusting.
Want more? Browse the store for more prompt packs.