Repository navigation
Use evals/results and honor Vally results configuration - #14
Merged
Merged
Conversation
Save native eval output under evals/results and resolve report defaults with Vally's project configuration loader, retaining explicit path overrides and native fallback behavior. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com>
There was a problem hiding this comment.
🟢 Approval recommended
No active review issues are reported for the changes aligning eval output and report discovery with project configuration.
No open issues 🎉
What changed in this PR
Aligns saved eval output and report discovery around Vally’s project configuration, using evals/results/ without moving historical results or changing plugin versions.
Changes:
- Configures and ignores
evals/results/while retaining the legacy ignore. - Makes report loading asynchronous and configuration-aware, preserving explicit paths and Vally’s default while surfacing malformed configuration.
- Adds regression coverage and updates eval documentation.
| File | Description |
|---|---|
docs/checks-and-evals.md |
Documents the configured results directory and explicit-path override. |
.vally.yaml |
Sets evals/results and wraps setup commands without changing their behavior. |
.gitignore |
Ignores the new results directory alongside the legacy location. |
.github/scripts/eval-report.test.mjs |
Covers configuration, fallback, overrides, CLI behavior, and malformed configuration. |
.github/scripts/eval-report.mjs |
Uses Vally configuration for default result discovery and awaits result loading. |
Review effort: Balanced (auto)
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.
What changed
Save native Vally eval output under
evals/results/and have the report helper follow the same project setting instead of independently assumingvally-results/..vally.yaml'spaths.resultsand ignore the new output directory while preserving the legacy ignore.@microsoft/vally/configto resolve the report's default results directory. Explicit result paths still take precedence, Vally's unconfigured fallback still works, and malformed project configuration is reported instead of silently ignored.Checks and boundaries
The report tests pass, including configured-vs-legacy selection, the real report CLI, explicit overrides and malformed configuration. Skill/eval lint and diff checks pass.
This is separate from #13's skill/grader fixes and the before/after workflow. No plugin versions, model defaults, sandbox settings or remote pins changed. No paid eval was run and no historical results were moved or committed.