Sign in
sources

Download crawl source raw file

Downloads the file content stored from crawl. Only crawl source IDs returned as content.source_id from /search, GET /contents/{id}, or POST /completions citations are valid. Local file source IDs accepted by GET /sources/{source_id} return 404 here.

The endpoint checks tenant ownership of the crawl source and returns the entire stored file. It does not apply per-chunk authorization policies; applications using enforce authorization should decide whether to expose raw files to each principal.

Required scope: `knowledge:read`

get/sources/{source_id}/raw

Path parameters

source_idstring uuid required

Crawl source ID returned in content.source_id

Query parameters

crawl_idstring uuid

Parent crawl ID. When supplied, the raw file is resolved through the crawl manifest without scanning the global crawled_files table. Omission uses the deprecated fallback.

Response

Stored crawl file content

Changes

Changed in 1 of the 71 revisions of this API.1

Of the 71 revisions, 18 have no diff computed.