Critique AI outputs and tag failures

Founder-testedBest with ClaudeReviews & critiquesby Startup CEO and Founder

When to use it

Evaluating two answers from an AI agent to improve it.

Here are [TWO] responses from [OUR AGENT] to the same input: [INPUT]. Which is better and why?

Flag any fabricated fact, wrong number, or missed option. Tag each failure by type (fabricated, lost, mis-bound, unconverted) and say what the ideal response would do.

Tune for my model →

0 copies · fill in the [PLACEHOLDERS] before sending. Never paste real customer data into an AI tool you don't control.

Next step: Try it in your AI assistant

Affiliate linkChatGPTPaste your tuned prompt and go Affiliate linkClaudeGreat for long, structured documents

Some links are affiliate or sponsored: we may earn a commission at no cost to you. Sponsor this spot →

More in Reviews & critiques

Founder-testedClaudeReviews & critiques

Compare two documents clause by clauseChecking a filled-in contract or proposal against its template.

▲ 0⧉ 0by Startup CEO and Founder

Founder-testedClaudeReviews & critiques

Feasibility and go/no-go for an opportunityDeciding whether to bid on a tender or big opportunity.

▲ 0⧉ 0by Startup CEO and Founder

Something wrong with this prompt? Contact us or submit a better one.