Skip to content

Improve review coverage and use runner settings - #4

Merged
dkropachev merged 1 commit into
mainfrom
improve-review-coverage
Oct 20, 2026
Merged

dkropachev merged 1 commit into
mainfrom
improve-review-coverage

Conversation

@dkropachev

Copy link
Copy Markdown
Contributor

Issue

Codex can discard concrete maintenance risks, and large combined instruction and diff reads can leave evidence unexamined. Model-specific reviewer effort rules can also conflict with the host or runner.

Fix

Share evidence criteria across Claude and Codex, inspect changed files before deep verification, recover omitted output, and reconcile every candidate explicitly. Preserve supported minor findings and disclose material coverage gaps in concise reports. Accept frozen PR metadata for offline Claude reviews and synchronize the plugin version to 1.0.4.

Decisions

  • Choice: The host or runner controls reviewer models and reasoning effort. Reason: the skill must not override the runtime configuration.

Share evidence criteria across hosts so supported maintenance risks and
minor findings remain visible. Inspect changed files before deep checks,
recover truncated output, and reconcile every candidate explicitly.

Leave reviewer models and reasoning effort to the host or runner and
synchronize the plugin manifests and catalog at version 1.0.4.
@dkropachev
dkropachev merged commit eddb395 into main Oct 20, 2026
1 check passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant