Does this skill review the report or the actual code?

Unlike basic agents, the Reviewer independently reads the actual code in the developer's worktree to verify that implementation matches the spec, regardless of what the developer's report claims.

When is the best time to trigger the Reviewer skill?

It should be activated after a developer agent reports task completion and the tester validates that all tests pass, but strictly before any code is merged into the main branch.

What happens if the Reviewer finds issues?

It generates a detailed report with specific file and line references, categorized by severity (Critical, Important, Minor), and provides clear fix instructions for the developer agent to address.

How does the two-stage review process work?

The agent first conducts a 'Spec Compliance' stage to ensure all requirements are met and no scope creep occurred. Only after this passes does it move to 'Code Quality' to evaluate patterns, naming, security, and performance.

Two-Stage AI Code Reviewer

Name: Two-Stage AI Code Reviewer
Author: johnhanks1

byjohnhanks1

0•

协作工具

Automates rigorous code reviews by validating specification compliance followed by comprehensive quality assessments directly within the developer's worktree.

The Reviewer skill serves as a high-integrity quality gate for AI-driven development workflows. It operates on a strict two-stage protocol: first, it independently verifies that the implementation aligns perfectly with the original task specifications and acceptance criteria; second, it evaluates the code for technical excellence, including patterns, performance, security, and maintainability. By reviewing directly in the developer's worktree before merging, it ensures that only spec-compliant, high-quality code reaches your main branch while providing line-specific, actionable feedback for any necessary fixes.

主要功能

01Independent verification of local worktree code before merging

02Standardized approval signals for CI/CD integration

03Strict two-stage review process (Spec Compliance and Code Quality)

04Line-by-line requirement tracking and gap analysis

05Comprehensive technical evaluation of patterns, security, and performance

060 GitHub stars

使用场景

01Ensuring total requirement coverage in complex software tasks

02Enforcing coding standards and preventing scope creep in team environments

03Automating quality control for AI-generated feature implementations

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add johnhanks1/vim-setup reviewer

For use in Claude.ai and ChatGPT