How does the code-reviewer subagent work?

The skill dispatches a subagent via the Task tool using specific git SHAs and requirements to provide an objective, systematic evaluation of the implemented work.

What are verification gates in this skill?

Verification gates are strict protocols requiring fresh, verifiable command-line evidence—such as passing tests or successful builds—before any status or completion claim can be made.

When is the best time to trigger this skill?

Use it when receiving feedback from any source, after completing any major task or feature, and immediately before making any statement that work is finished or bugs are fixed.

How does this skill prevent 'lying' agents?

It implements an 'Iron Law' that forbids claiming success without fresh evidence, forcing the agent to run verification commands and read the output before reporting status.

Why does this skill discourage 'performative agreement'?

It prioritizes technical correctness over social comfort, ensuring developers fully understand and verify feedback before implementation rather than defaulting to polite but unverified acceptance.

Code Review & Verification

Name: Code Review & Verification
Author: rafaelcalleja

byrafaelcalleja

•

Collaboration Tools

Enforces technical rigor and evidence-based validation throughout the code review and task completion lifecycle.

This skill transforms the code review process by replacing performative agreement with technical validation and strict verification gates. It provides structured protocols for receiving feedback, systematically requesting reviews via subagents, and enforcing an 'Iron Law' that requires fresh evidence before any completion claims can be made. By using this skill, developers ensure that every change is technically sound, every claim is backed by command output, and every piece of feedback is evaluated with professional skepticism rather than social comfort, making it essential for high-integrity subagent-driven development.

Key Features

01Automates systematic review requests using specialized code-reviewer subagents

02Enforces line-by-line requirement checklists before merging or pushing code

03Implements mandatory verification gates requiring fresh evidence for completion claims

04Eliminates performative agreement in favor of technical correctness and rigor

051 GitHub stars

06Provides structured decision trees for handling unclear or questionable feedback

Use Cases

01Processing complex external feedback without compromising existing architectural decisions

02Validating major feature deployments or refactors before merging to the main branch

03Preventing false completion claims in subagent-driven development environments

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add rafaelcalleja/claude-market-place code-review

For use in Claude.ai and ChatGPT

Download Skill