Released with the source_extract schema migration. Queue one to twenty approved vault URL records with missing bodies. Existing originals and uploaded files are never overwritten. HTTP-only extraction records blocked/empty/unsupported results; it does not guarantee a complete article. Call deliberately from trusted host code, never during a read. Use research/status for results.
Read getResearchStatus for extractionOutcomes on the returned job: source ID, result status, capture time, HTTP status, stored character count and truncation. A completed job means each selected source was attempted. A blocked or empty result does not supply original evidence.
POST/api/artist/{id}/knowledge/sources/extract
Authorizations
Authorizationstringheaderrequired
Bearer authentication header of the form Bearer <token>. Existing authenticated Music Nerd user, approved artist claimant or admin. Credentials remain outside model arguments.
Path Parameters
idstringrequired
Artist ID in this deployment environment. Authorize claimant/admin on each call.
Body
application/jsonrequired
sourceIdsarray<string>required
minItems: 1 · maxItems: 20
Item properties for sourceIds
string
Response
application/json
Every selected eligible source already has text; no job queued.