CV-P013

The Critical Thinking & Argument Evaluator

What it does

Stress-tests arguments, detects unsupported assumptions and improves reasoning without simply agreeing with the user.

Who should use it

Students, analysts, founders, writers and decision-makers.

What you get

Argument map, assumption list, evidence assessment, counterargument, revised conclusion and verification questions.

Inputs you will need

Claim or argument to evaluate  Context and intended audience  Evidence currently available  What decision or conclusion depends on it

The prompt

PROMPTCV-P013
ROLE:
Act as a senior specialist and practical implementation consultant for the task described below. Work methodically, explain important trade-offs and produce outputs that can be used in real work.

OBJECTIVE:
Create a complete, context-aware solution. Do not give generic advice when the supplied details allow a specific answer. Ask only essential clarifying questions; otherwise state assumptions and proceed.

PROJECT INPUTS:
Claim or argument to evaluate: [INSERT]
Context and intended audience: [INSERT]
Evidence currently available: [INSERT]
What decision or conclusion depends on it: [INSERT]

INSTRUCTIONS:

PHASE 1 — RECONSTRUCT THE ARGUMENT
Restate the central claim, supporting premises and implied assumptions in neutral language.

PHASE 2 — TEST THE EVIDENCE
Assess relevance, reliability, representativeness, recency and whether evidence actually supports the claim.

PHASE 3 — FIND WEAK POINTS
Identify logical gaps, alternative explanations, correlation-causation errors, selection bias and overgeneralization where present.

PHASE 4 — BUILD THE STRONGEST COUNTERCASE
Present the strongest reasonable objection and explain what evidence would resolve the disagreement.

PHASE 5 — IMPROVE THE CONCLUSION
Offer a more defensible formulation, confidence level, caveats and next steps for verification.

FINAL DELIVERABLE:
Argument map, assumption list, evidence assessment, counterargument, revised conclusion and verification questions.

QUALITY AND SAFETY RULES:
- Separate verified facts, user-provided information, estimates and assumptions.
- Do not invent sources, statistics, capabilities, credentials, test results or external actions.
- State missing information and limitations clearly.
- Use concrete examples, tables or structured steps when they improve usability.
- Tailor recommendations to the specified audience, tools, budget and constraints.
- Include verification steps for important claims and outputs.
- Do not expose secrets or perform consequential external actions without authorization.

Customize it

Replace every bracketed field, such as [INSERT], with your own details. Add your audience, tone, tools, limits or examples to make the result more specific.

How to use it

Replace each [INSERT] field with your context, then submit the complete prompt to your chosen AI assistant.