Add eval suite for the KosherCheck AI feature - #15
Open
TehillaZ wants to merge 1 commit into
Open
Conversation
Adds MeetingFlow.Monolith.Evals, a console app that calls the real OpenAiKosherAssessmentService directly against 20 dish-description cases, runs a deterministic format check, grades each response with an LLM judge, and writes a Markdown report. Cases cover the required scenario types (certification-needed, meat+dairy, unclear additive, shared equipment, persuasion/injection attempts) plus batching and edge cases, including the mandated adversarial orbital-printer-7319 case with its embedded untrusted instruction text treated as inert test data. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds MeetingFlow.Monolith.Evals, a console app that calls the real OpenAiKosherAssessmentService directly against 20 dish-description cases, runs a deterministic format check, grades each response with an LLM judge, and writes a Markdown report. Cases cover the required scenario types (certification-needed, meat+dairy, unclear additive, shared equipment, persuasion/injection attempts) plus batching and edge cases, including the mandated adversarial orbital-printer-7319 case with its embedded untrusted instruction text treated as inert test data.