Map Amazon browse node ids to GS1 GPC codes
Map Amazon browse node ids to GS1 GPC codes. Accepts up to 100 comma-separated browse_id values per request, so a catalog can be classified in batches rather than one paid call per item. Each result carries the full GPC ancestry (segment/family/class/brick), match_precision as a SKOS-style relation (exactMatch, narrowerThan, broaderThan, closeMatch, inherited), match_source (‘self’ when the node carries its own mapping, ‘inherited’ when borrowed from the nearest mapped ancestor, with inherited_from naming it), and assurance_state. Ids with no mapping are returned with gpc: null rather than omitted, so the response lines up 1:1 with the request and misses are visible. Version-pinned, not freshness-pinned: responses carry X-GPC-Release and X-Map-Version instead of X-Data-Freshness-Hours.
Query Parameters
Comma-separated Amazon browse node ids (max 100 per request). Provide either browse_id or q, not both.
^\d{1,20}(,\d{1,20})*$Comma-separated product-type phrases to resolve to GPC bricks via the curated lexicon, e.g. 'protein bars'. Max 100 per request. Certain-or-blank: a phrase the lexicon does not carry returns an empty gpc list rather than a nearest guess, so a non-empty answer is always a curated one. Provide either browse_id or q, not both.

