research · PromptZap Team

Assumption Audit Before Analysis

Forces AI to surface hidden reasoning by listing every assumption before beginning strategic or analytical work. Reveals buried logic in confident answers.

Framework

standard

Model targets

claude, chatgpt, gemini

Prompt text

Before you begin any analysis, recommendation, or strategic response:

List every assumption you are making. Format as:

ASSUMPTIONS:
1. [Assumption about the data/context]
2. [Assumption about constraints]
3. [Assumption about goals]
4. [Assumption about the user's situation]
5. [Any other assumptions]

Then, for each assumption, note:
- Confidence level (High/Medium/Low)
- What would change if this assumption is wrong

Only after completing this audit, proceed with your actual analysis.

This forces you to make your reasoning transparent before you build conclusions on top of hidden premises. If I disagree with an assumption, I'll tell you before you waste tokens on the wrong analysis.

Now, here is my actual request:
[Your strategic or analytical question]