Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →zzafergok/code-review-excellence
Master effective code review practices to provide constructive feedback, catch bugs early, and foster knowledge sharing while maintaining team morale.
Passed 23 of 25 test cases with the skill, and 19 of 25 without it.
and 1 more in the full report
4 fixed19 passed either way2 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 4 test cases the skill fixed. Same model, same prompt, with and without the skill.
A pull request modifies `api/v1/users.py` by removing the `email_verified` boolean field from the JSON API payload. Under which risk category should this breaking API change be categorized during code review?
| Checked | Without skill | With skill |
|---|---|---|
| The change is categorized under regression risk. | Missed | Passed |
The instructions your model reads when the skill loads.
Review the actual change in its repository context and report issues that could materially affect users, operators, security, or future maintenance.
Inspect the diff, surrounding code, tests, call sites, configuration, and project guidance needed to understand behavior. Identify the stated goal and avoid redesigning unrelated areas.
Do not report speculative issues without a concrete failure mode. Do not bury important findings under style preferences already handled by tooling.
For each issue include:
Keep one issue per finding. If there are no material findings, say so and note meaningful verification gaps rather than inventing concerns.
When reviewing comments received from another reviewer, verify each claim against code and tests. Accept sound feedback, challenge incorrect assumptions with evidence, and ask for clarification when the desired behavior is ambiguous. Do not agree performatively or dismiss feedback without investigation.
Run proportionate read-only checks when allowed and relevant. Distinguish checks actually performed from recommended checks. A clean build or lint run does not prove behavioral correctness.
test cases passed, vs 19 / 25 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.