feat(traces): give the trace evaluator data-fair data-exploration tools#37
Merged
Conversation
Read-only data-exploration tools for the trace evaluator, reusing @data-fair/agent-tools-data-fair here and calling the same-origin data-fair API scoped to the conversation's owner account. Includes a raw metadata tool for metadata-quality and tool-coverage evaluation. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…field-values tools Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Also covers the search_data next-pagination branch with a test. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Merge buildEvaluatorDataTools into the evaluator localTools and thread the conversation owner's department through TraceReview -> EvaluatorChat. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…ontract Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
# Conflicts: # ui/src/components/EvaluatorChat.vue # ui/src/components/TraceReview.vue
# Conflicts: # ui/src/components/TraceReview.vue
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 read-only data-exploration tools to the admin trace evaluator so it can check a reviewed session against the actual data instead of judging from the trace text alone.
evaluator-data-tools.tsbuildslist_datasets,describe_dataset,get_dataset_schema,search_data,aggregate_data,calculate_metric,get_field_values(reused from@data-fair/agent-tools-data-fair) plus an evaluator-specificget_dataset_metadata_rawfor the full untrimmed metadata.owner=type:id[:department]), so a superadmin reviewing account X explores X's data. They register always and degrade gracefully when data-fair is unreachable.EvaluatorChat's tools;departmentthreaded throughTraceReview; evaluator prompt updated.Why: so the evaluator, when opened in a data-fair context, has data-exploration tools like the normal agent and can extend its evaluation to the real data (e.g. missing dataset descriptions, schema quality, whether a search would return rows).
Heads-up: superadmin cross-account exploration relies on an active
adminModesession (no data-fair change);publicationSitescoping is deferred (not captured in traces). Adds the@data-fair/agent-tools-data-fairUI dependency.