agent_content

Bounded anonymous reads: search, resource/+id, related/+id and schema. Requests accept binary-keyed maps. Public listings exclude noindex resources. Direct reads retain public noindex content. No totals or private properties are returned.

Search accepts text, cat, language, page, pagelen, sort (id/newest/oldest/modified), and inclusive UTC publication_after/publication_before dates (YYYY-MM-DD). Related reads accept direction (in/out) and an optional public predicate ID/name. Schema reads accept kind (category/predicate) and describe public category parents and bounded predicate category constraints. They are vocabulary, not ACL rules. All reads default to the site language, use a five-second budget, and cap pages at 100 candidates. Follow next_url even after an empty filtered page.

Edit on GitHub