Best agent skills for code review
Review agent changes against intent, repository standards, and production quality risks.
How to choose
Matt separates standards and specification fidelity so one concern does not hide the other. Superpowers treats review as a two-way discipline with a context-isolated reviewer, Pstack brings multi-model adversarial review, and Addy expands the review surface to maintainability, tests, security, performance, and simplification.
Code Review
Reviews a diff independently against repository standards and the originating specification.
Codebase Design
Designs deep modules with small interfaces, clean seams, and testable behavior.
Our take: Codebase Design gives agents a compact architectural vocabulary centered on deep modules, narrow interfaces, locality, and testable seams. It improves design conversations without pretending that a generic architecture pattern fits every repository.
Requesting Code Review
Sends a change to a context-isolated reviewer before it can cause cascading problems.
Receiving Code Review
Evaluates review feedback with technical rigor instead of performative agreement or blind implementation.
Interrogate
Has several different models try to break a diff, including a strict code-quality lens.
No Comments
Strips comments before review, runs the Comment Sicko reviewer, and fixes accepted findings.
Code Review and Quality
Reviews changes across correctness, maintainability, testing, security, and performance.
Our take: A broad pre-merge review that treats correctness as only one axis. Its value is the systematic inclusion of maintainability, test quality, security, and performance before a change enters main.
Code Simplification
Reduces accidental complexity without changing externally visible behavior.
Doubt-Driven Development
Subjects non-trivial decisions to adversarial fresh-context review before they stand.