Record a positive or negative verdict on a signal. Feedback trains the scoring model over time.
positive — this signal was useful; reinforce scoring rules that surfaced itnegative — this signal was noise; penalize scoring rules that surfaced itnegative feedback tells the system the signal was actually incorrect. An agent filtering by confidence threshold should dismiss without feedback; a human marking “this isn’t relevant” should submit negative feedback.Body shape. JSON body with exactly one field: feedback: "positive" | "negative". Any other value returns 400.Overwrite semantics. Calling negative then positive on the same signal leaves it as positive. There’s no “clear feedback” — if you need to retract, submit the opposite.This is the public partner API — rate-limited at 100 req/min. Use this rather than the internal /api/signals/feed/{id}/feedback endpoint.Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Signal 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)$Overwrites any previous feedback on this signal — there is no history, just the latest value.
positive — this signal was useful; negative — this signal was noise. Feedback improves scoring over time.
positive, negative Feedback recorded.
true