fix(results): tighten result artifact contract - #1645
Merged
Conversation
christso
force-pushed
the
results-artifact-contract-tighten
branch
from
July 4, 2026 08:21
5d16c80 to
9e1ec27
Compare
Deploying agentv with
|
| Latest commit: |
9e1ec27
|
| Status: | ✅ Deploy successful! |
| Preview URL: | https://c9956720.agentv.pages.dev |
| Branch Preview URL: | https://results-artifact-contract-ti.agentv.pages.dev |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
AgentV result bundles now expose the current tags-first contract without the leftover experiment/config compatibility fields from the first cleanup pass. Run summaries keep
experimentas run metadata and keepruntime_sourcefocused on runtime provenance:schema_version,config_source,eval_files, and optionalwrapper_eval_file.This removes the redundant public fields called out in av-xz5i: root
summary.jsonpass_at_k, duplicated experiment namespace/source metadata,runtime_source.kind,source_eval_files, and the public.internal/run-config.json/experiment_configpath. Dashboard, serve, report, docs, and artifact-writer tests now consume that slimmer shape directly.Comparable-framework check: Promptfoo and Margin keep machine duration fields in milliseconds (
durationMs/latencyMs,runtime_ms/run_ms), while Vercel agent-eval converts agent milliseconds into seconds for result duration display. AgentV keeps machine-readable durations in ms and leaves human seconds to renderers.Related: av-xz5i
Validation
bun run buildbun run typecheckbun test apps/cli/test/commands/eval/artifact-writer.test.ts apps/cli/test/commands/eval/tag-filtering.test.tsbun test apps/cli/test/commands/results/report.test.ts apps/cli/test/commands/results/serve.test.tsbun test apps/dashboard/src/lib/run-detail-context.test.ts apps/dashboard/src/components/RunList.mobile.spec.tsxbun test apps/cli/test/eval.integration.test.tsbun apps/cli/src/cli.ts eval run examples/features/rubric/evals/operators.eval.yaml --target azure --grader-target azure --test-id finance-summary-operator-guards --workers 1 --experiment av-xz5i-dogfood --output .agentv/results/av-xz5i-contract-dogfood-azure-20260704 --no-cachepassed 1/1 with a live Azure target and Azure LLM rubric graderEvidence
Private evidence branch:
EntityProcess/agentv-private@evidence/av-xz5i-result-artifact-contract-2026-07-04(862965c). It includes the passed dogfood run bundle, source eval, and contract inspection notes showing the removed fields are absent from the generated public artifact bundle.