Skip to main content

Crawl per-content fabric metadata into 'content' documents

POST 
/spaces/:qid/update_content_metadata

Runs a lightweight fabric crawl over the space's collection (or a subset, when qids is supplied) using the space's fabric.indexer config. Each crawled content object becomes one Document with track="content", title from public.name, and text composed of every configured field value joined by newlines. Existing same-qid vectors in the space are linked to the new document via the vector_documents mapping. Crawling runs before any writes, so failures leave the database untouched. Requires content access for the space qid and the supplied auth token to be valid against each crawled content object.

Request​

Responses​

OK