When is it mandatory to use the code review skill?

It is highly recommended after completing each task in subagent-driven development, after finishing a major feature, and always before merging into the main branch.

How does the subagent know which code to review?

The skill uses Git SHAs (BASE_SHA and HEAD_SHA) to define the specific range of commits that were modified during the task.

What happens if the AI reviewer identifies an issue?

Feedback is categorized by severity. Critical and Important issues should be fixed immediately before proceeding, while Minor issues can be noted for later.

Can I disagree with the AI's code review feedback?

Yes. If the reviewer is incorrect, you should push back with technical reasoning or provide tests that prove the implementation works as intended.

Code Review Assistant

Name: Code Review Assistant
Author: axiomantic

byaxiomantic

Collaboration Tools

Automates the dispatch of subagents to review code changes and verify requirement alignment before merging.

About

This skill provides a structured framework for requesting automated code reviews within Claude Code. By leveraging git commit SHAs and specialized subagent templates, it helps developers identify critical issues, technical debt, and requirement gaps early in the development cycle. It is particularly effective for subagent-driven development workflows, ensuring that every task is verified by a 'fresh set of eyes' before proceeding to the next step or merging into the main branch.

Key Features

Support for both task-based and ad-hoc development reviews
0 GitHub stars
Template-driven review requests for consistent quality checks
Automated subagent dispatching for targeted code analysis
Git-integrated commit comparison using BASE and HEAD SHAs
Structured feedback categorization (Critical, Important, Minor)

Use Cases

Gaining a fresh perspective and debugging help when stuck on complex code
Catching logic errors between individual tasks in a complex development plan
Verifying major feature implementations before final merge to main

About

Key Features

Support for both task-based and ad-hoc development reviews
0 GitHub stars
Template-driven review requests for consistent quality checks
Automated subagent dispatching for targeted code analysis
Git-integrated commit comparison using BASE and HEAD SHAs
Structured feedback categorization (Critical, Important, Minor)

Use Cases

Gaining a fresh perspective and debugging help when stuck on complex code
Catching logic errors between individual tasks in a complex development plan
Verifying major feature implementations before final merge to main