Read the per-source configuration rows for a signal type — which sources are enabled, with any source-level keyword and prompt overrides.
signal_type_sources view for a signal type — one entry per registered source (news, reddit, hacker_news, x_twitter, linkedin_posts, linkedin_jobs, g2_capterra), whether enabled or not.
Each entry carries:
enabled — does this source feed into this signal type?keywords_extra — additional keywords used ONLY when searching this source (in addition to the type-level keywords)prompt_override — custom classification prompt replacing the type’s default for this source. null = use the type’s default.PATCH /api/v1/signals/types/{id}/sources/{sourceId}.
enabled === true if you only want the active scope.Relationship with type-level sources. The type-level sources array (from GET /api/v1/signals/types/{id}) is a quick allowlist view; this endpoint is the full per-source settings. The two are kept in sync server-side when PATCH /api/v1/signals/types/{id} mutates sources.Extra keywords narrow, don’t replace. keywords_extra on a source entry is ADDITIVE — the scanner uses the union of type-level keywords + the source’s extras. To restrict a source to a specific keyword set, combine the PATCH with type-level keyword tuning.This is the public partner API — rate-limited at 100 req/min. Use this rather than the internal /api/signals/types/{id}/sources endpoint.Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Signal type UUID.
^([0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[1-8][0-9a-fA-F]{3}-[89abAB][0-9a-fA-F]{3}-[0-9a-fA-F]{12}|00000000-0000-0000-0000-000000000000|ffffffff-ffff-ffff-ffff-ffffffffffff)$Array of source-level configs for this signal type.
Per-source settings for this signal type, one row per registered source.