Write a reusable prompt with replaceable inputs. Submit your prompt and one test output for this example.
Task: Use AI to build and challenge a simple vendor scorecard.
Example input: Three fictional tools receive scores from 1 to 5, where higher is better. Atlas: cost 5, usability 3, integration 2, support 3. Birch: cost 3, usability 4, integration 5, support 4. Cedar: cost 4, usability 5, integration 3, support 2. Baseline weights are cost 30%, usability 30%, integration 25%, and support 15%. A second stakeholder wants integration weighted at 40%; reduce cost and usability to 25% each and support to 10%. These scores are exercise inputs, not market research.
Constraints:
- Show the weighted calculation for every option under both weight sets.
- Do not treat small score differences as objective certainty.
- Name one important consideration the scorecard does not capture.
Required output from your prompt: Produce two ranked scorecards, a conditional recommendation, and a short explanation of whether the preference change matters.
Success checks:
- Each weight set totals 100%.
- The weighted totals can be reproduced from the inputs.
- The recommendation matches the calculated rankings.
- The limitations distinguish assumed scores from verified product evidence.