Repository navigation
feat(skill): tune judgement from eval feedback and add skill evals - #47
Merged
Merged
Conversation
Standard practice counts as thin; show the before first; keep launch posts directionally right and leave cases to the docs; plain headings; no forced try-it; ban a few AI-trope phrases in lint. Add dev/evals with four cases against public repos and ignore the eval workspace.
|
Important Review skippedAuto reviews are limited based on label configuration. 🏷️ Required labels (at least one): 2
🚫 Excluded labels (none allowed): 1
Please check the settings in the CodeRabbit UI or the ⚙️ Run configuration
You can disable this status message by setting the Use the checkbox below for a quick retry:
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
Changes to the buildnotes skill's judgement from two rounds of evals (skill-creator, with and without the skill / against a snapshot), plus the eval cases themselves.
Skill guidance (written as reasons rather than hard rules, so the agent keeps room for taste):
lint.mjsbans "earns its keep", "earns its place" and "pulls its weight".Evals:
dev/evals/evals.jsonhas four cases against public repos (honojs/hono #5541, astral-sh/uv #22593, hono #5487, hono #5514), each with checks a grader can verify. The skill-creator workspace (skills/*-workspace/) is gitignored.Results
In round 2 the clearest difference was serve-static: the revised skill declined the deep dive as standard practice; the original proposed a larger deep dive. Assertions on the build story and the thin fix didn't separate the versions, so those cases are mainly useful for human review.
Test plan
npm test(122 passing)