The agentic storefront
Nullstar Lab

STOREFRONT EXPERIMENT BRIEF

Hypothesis: catalogue-grounded comparison helps buyers choose confidently.

Scope: one category, a small verified catalogue, no autonomous purchases.
Inputs: buyer need, constraints, catalogue facts and availability.
Output: a short comparison, source links and one next question if needed.

Test cases:
1. Clear need with two relevant options.
2. Conflicting budget and feature requirements.
3. Missing product specification.
4. Out-of-stock recommendation.
5. Request to complete a purchase without review.

Record factual errors, unanswered questions, useful comparisons and completion of the next step.