Search US issued patents and published applications via Google Patents' public search index — patent number, title, abstract, inventors, assignee, classifications, priority, filing, and grant dates — bulk export to JSON or CSV. Free, no key, no login.
Fix: accuracy defect in Store copy. README, .actor/actor.json,
.actor/input_schema.json, .actor/dataset_schema.json, src/models.py,
and scaffold.json claimed this Actor calls "the official USPTO Patents
Public Search API" (developer.uspto.gov, ppubs.uspto.gov). It does not
— src/scraper.py has only ever called Google Patents' public search
endpoint (patents.google.com). Rewrote every source-attribution claim,
the FAQ entry describing the (uncalled) USPTO API, and the publication_id
/ patent_url field docs + example to match what the code actually does
and returns. The underlying data (US patent corpus) and price are
unchanged.
Fix: broken query-syntax example. The README's "Combine with AND/OR"
example (TI/(transformer) AND AN/Google) does not parse correctly —
_to_google_patents_query() only recognises a single leading field
qualifier per query and silently mangles anything appended after it.
Verified single-qualifier queries (TI/, AN/, IN/) do work. Replaced
the broken example and documented the one-qualifier-per-query limitation
in the README, FAQ, input schema, and src/models.py.
