Repository navigation
feat(skills): include behavior and impact in native and delegated reviews - #2029
Open
lizhengfeng101 wants to merge 2 commits into
Open
lizhengfeng101 wants to merge 2 commits into
lizhengfeng101 wants to merge 2 commits into
Conversation
Contributor
|
✅ OpenCodeReview: Review skipped: no items were selected. |
This branch has not been deployed
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
The native skill now asks OCR v1.14.0 or later for behavior and impact analysis on fresh commit and range reviews by default. A request to review a commit is enough; the caller does not have to know the flag. Workspace reviews, previews, resumes, SARIF output, and explicit requests to skip analysis keep their existing invocation. The CLI default does not change.
The standalone skill and its plugin mirror distinguish Review findings and coverage from Analysis execution and the model's assessment of completeness. This keeps a partial investigation from being presented as a clean review. Analysis has a separate token budget, so eligible reviews can incur additional model usage.
In delegated reviews, the host considers the same user-facing question while reading the diff, including workspace changes. There is no new flag, separate native run, or OCR provider requirement. The host reports concrete impacts separately from defect counts and omits the section when it has nothing material to add.
Validation
make checkandmake testpassed (28 packages).deletedand available on the old side, the explicitly excluded path remaineduser_exclude, and neither counted as reviewable.ocr reviewruns selected zero files: the changed Markdown files were excluded asunsupported_ext. The documented flags and result fields were checked against the v1.14.0 implementation separately.Related Issues
Refs #2024 (part of #2023); builds on #1937. The default-on behavior for eligible skill invocations supersedes the opt-in wording in the original issue; snapshot management is out of scope.
AI Disclosure
OpenAI Codex (
gpt-6.2-astra) was used for the skill changes, validation, and PR preparation.