feat(worker): detect and call out model refusals (#30)#374
Draft
tuliren wants to merge 1 commit into
Draft
Conversation
Adds a tested refusal detector (packages/lib/refusal-detection) and wires it
into the worker: when a model declines to answer ("I can't help with that",
"I must decline"), the run is logged, reported to telemetry as a
prompt_refused event, and excluded from brand/competitor mention counting so
refusals don't masquerade as visibility. Surfacing refusal rates in the
dashboard UI is a follow-up.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
@tuliren is attempting to deploy a commit to the Blue Whale Labs Team on Vercel. A member of the Team first needs to authorize it. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Detects when a model refuses to answer ("I can't help with that", "I must decline") and explicitly calls it out.
Changes
packages/lib/src/refusal-detection.ts:detectRefusal()(opening-segment + short-response scan, curly-apostrophe tolerant).prompt_refusedevent, and excluded from brand/competitor mention counting so they don't masquerade as visibility.Testing
refusal-detection.test.ts; workertsc --noEmitpasses.Note
This covers detection + exclusion + telemetry. Surfacing refusal rates in the dashboard UI (likely a schema column) would be a sensible follow-up.
Refs #30
🤖 Generated with Claude Code