What columns work best?
Product name is the minimum. Brand, category, model, UPC, SKU, lot code, and manufacture date make the result more useful.
CSV recall-risk scan
Free recall-risk CSV scan for resale inventory.
Send up to 100 product rows. Get a review-ready CSV with possible matches, confidence scores, matched fields, source links, and suggested review actions.
Best fit
Built for messy catalog rows before a full integration.
This is the fastest way to validate whether recall screening is useful for your intake, marketplace, or inventory workflow.
- Rows
- Up to 100
- Best fields
- Product, brand, model, UPC, lot, category
- Output
- Status, confidence, matched fields, source URL
01
Send a small export from listings, POS, inventory, or catalog software.
02
Rows are normalized and checked against connected recall sources.
03
You get a review queue, not a claim that items are safe.
Returned CSV
What the review file should answer.
The scan is designed to separate normal workflow rows from rows that deserve a human look, lot-code check, or official-source verification.
example output
review queue
row_id,product_name,brand,status,top_confidence,matched_fields,source,recommended_action,source_url
A001,Graco stroller,Graco,no_match,0.12,"",,normal_review_flow,""
A002,Fisher-Price swing,Fisher-Price,possible_match,0.84,"brand,product_type",CPSC,manual_review_required,https://www.cpsc.gov/Recalls
A003,Jif peanut butter 16 oz,Jif,needs_lot_verification,0.88,"brand,product_name",openFDA,verify_lot_code,https://api.fda.gov/food/enforcement.json
A no_match result means no matching recall was found in
connected sources. It does not mean the product is safe or not
recalled.
FAQ
What to expect from the free scan.
What happens after I submit?
The request goes into the support queue. We review the sample, run the scan when it fits the pilot, and reply by email.
Can this replace official checks?
No. CatalogRecall API provides automated matching only. Verify possible matches against official recall notices before acting.
Can I integrate instead?
Yes. Use POST /v1/recalls/match for one row or
POST /v1/recalls/bulk-csv for API-level batch checks.