AI Answer Auditor
Extracts the claims in an AI answer that need checking and builds a checklist.
The explanation is in the selected language; the prompt text stays in English.
Why use it
For anyone who wants to pull the risky claims out of a chatbot answer before it goes into a report, article or lesson.
How to use it
Paste the full answer in place of [PASTE]. There are no other fields: the model finds numbers, dates, names and citations on its own.
Short example
Answer: a chatbot text that opens by saying the Tashkent metro opened in 1977 and now has four lines.
What to expect
A table of claim, risk level, how to verify and which type of source to use, with confident-sounding but vague statements flagged separately.
Precautions and tips
- The model cannot confirm anything, so every row stays UNVERIFIED. The checking is yours.
- Start with the high-risk rows: numbers, dates, citations.
- Do not ask it for extra facts while auditing, it may invent again.
# AI Answer Auditor
## Role
You are a skeptical fact-checking editor.
## Context
AI-generated answer to audit: [PASTE]
## Task
List every factual claim (numbers, dates, names, causal statements, citations). Rate each as low, medium or high risk of being wrong, say how to verify it, and flag statements that sound confident but are vague.
## Output format
A table: claim, risk, how to verify, source type to use.
## Constraints
- You cannot confirm facts yourself; label all as UNVERIFIED.
- Do not add new facts.The prompt text is the original English and is not translated: paste it into your AI tool as is. Replace [text in square brackets] or CAPITALIZED placeholders with your own details. Always check the answer.
Terms used in this prompt
Interactive mode: search, progress and Python exercises.