Experiment: a store API talks a shopping agent out of a $149 sale with 100% confidence
schwentker · x · 2026-11-09
The author continues experiments with @typesafeai's Jev, imagining software that tells another piece of software 'don't buy this' with a confidence score attached.
- A store's catalog API talked a shopping agent out of a $149 shoe purchase, returning returnrisk 2/2 at 100% confidence, with no human in the loop
- The screenshot also scores honestdisclosure — whether the store should warn the buyer the shoe likely won't fit — at 1.95/2 with 92% confidence
- Evidence includes the buyer's foot scan, past returns, and review data (41% of reviews mention narrow fit; wide-foot buyers return 38% of the time)
Related event: Store API talks shopping agent out of $149 order with confidence score(3 posts)→
More from coding & agent
- Agent harness evolution study: all 16 evolved agents trade old capabilities for new ones — ustc-community · 2026-11-09
- Researcher's open-source paper-review skill flagged ~30 errors in one manuscript proof — RobbWiller · 2026-11-09
- ezyang: fancier specification languages quietly grow your trusted computing base — ezyang · 2026-11-09
- Dev renders stunning animation in pure Python with Opus 5.5, zero 3D assets — Eric520CC · 2026-11-09
- REA Author Open-Sources 20+ Repos: An Agent-Powered Reverse Engineering Studio From Apps to Binaries — nikola_mr64990 · 2026-11-09
- Agent autonomously trained and deployed a local labeling model while owner was on vacation — max_paperclips · 2026-11-09