090°Open data
Open datasets, fully documented — searchable here, and readable by any LLM.
1 datasets
FDA recall enforcement intelligence (agent-curated)
Every FDA enforcement-report recall (food, drug, device) with a report_date in the trailing 30 days ending at the openFDA feed last_updated, from the official keyless openFDA enforcement APIs. Records are deduped on recall_number; reason_for_recall is classified into a closed 11-category taxonomy by ordered keyword rules; firms are canonicalized (suffix/location-tail stripping); distribution_pattern is parsed into a closed scope taxonomy; countries resolve to ISO alpha-3 via hub.normalize; UPC digit runs and embedded e-mails are redacted at parse time. Each recall carries a documented 0-100 recall_risk_score (classification severity + distribution scope + reason severity + ongoing status) with a deterministic recall_rank. Only factual recall metadata is stored — no copyrighted expression. Caveat: the feed lags the real enforcement calendar by ~10 days; report_date is the FDA publication date, not the recall initiation date. Primary key: recall_number. Cadence: monthly.
Once today's free allowance is used up, AI features can run on your own provider account.
Kept in this browser tab only (cleared when you close it) and sent with each AI request. Our servers use it for that request and never store or log it.