feat(evals): link Datadog experiments to runtime traces - #3302
Draft
mehulsonowal wants to merge 2 commits into
Draft
mehulsonowal wants to merge 2 commits into
mehulsonowal wants to merge 2 commits into
Conversation
Signed-off-by: mehul.sonowal <mehul.sonowal@datadoghq.com>
Signed-off-by: mehul.sonowal <mehul.sonowal@datadoghq.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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Eve evals can already observe the runtime trace contexts for every participating session, but the Datadog reporter drops those references when it submits Experiment rows. This change records a content-free, namespaced
eveRuntimeTraceLinksarray in each row's metadata so Datadog can resolve the canonical runtime traces at query time without changing trace IDs or parentage. Repeated contexts are deduplicated and a primary-session context wins when the same trace/span pair appears more than once.This is the producer PR for the cross-repository query-time trace-expansion work. It does not pre-create Experiment spans, copy prompts or outputs into link metadata, or change runtime instrumentation.
Related draft PRs:
Validation
pnpm --filter eve build:compiledpnpm --filter eve exec vitest run --config vitest.unit.config.ts src/evals/runner/reporters/datadog.test.ts --reporter=verbose— 14 tests passedpnpm exec oxlint packages/eve/src/evals/runner/reporters/datadog.ts packages/eve/src/evals/runner/reporters/datadog.test.ts— no warnings or errorspnpm --filter eve typechecktest-tuiand secret-backed E2E/Vercel jobs fail on this fork PR because GitHub withholdsAI_GATEWAY_API_KEYand Vercel credentials; the logs reportMODEL_CALL_FAILED: AI Gateway received no credentialsbefore exercising this change.Checklist
CONTRIBUTING.mdevepackagegit commit --signoff)