What is the implementation order for multiple feedback items?

Claude follows a prioritized sequence: first clarifying the unknown, then fixing blocking/security issues, followed by simple fixes, and finally complex refactoring, with individual testing at each step.

What happens if a human partner gives unclear instructions?

The skill triggers a 'STOP' protocol. Claude must identify exactly which items are unclear and ask for clarification before implementing any part of the feedback to avoid partial or incorrect solutions.

How does Claude handle feedback that seems incorrect?

Claude is instructed to provide technical pushback based on codebase reality, such as checking build targets, grep-ing for usage, or identifying potential regressions rather than simply following orders.

Why does this skill discourage saying 'Thanks' or 'You're right'?

The skill prioritizes 'actions over words.' By removing performative social cues, it ensures the focus remains entirely on technical verification and the actual code changes, preventing blind agreement with incorrect suggestions.

Code Review Reception

Name: Code Review Reception
Author: lifangda

bylifangda

•

Collaboration Tools

Establishes a rigorous technical protocol for Claude to evaluate, clarify, and implement code review feedback without performative agreement.

The Code Review Reception skill optimizes how Claude processes feedback by prioritizing technical accuracy over social politeness. It shifts the interaction model from blind implementation to a 'verify-first' approach, requiring Claude to restate requirements, check against codebase reality, and provide reasoned pushback when suggestions are technically unsound. This skill is particularly valuable for maintainers who want an AI collaborator that acts as a critical thinker, ensuring that every change—whether from a human partner or an external reviewer—is necessary, tested, and aligned with the YAGNI (You Ain't Gonna Need It) principle.

Key Features

01Mandatory clarification protocol for ambiguous or partial feedback

02Strict prohibition of performative agreement and unnecessary gratitude

03Multi-step verification workflow (Read, Understand, Verify, Evaluate, Respond, Implement)

04Differentiated handling for human partners versus external reviewers

05Automated YAGNI checks to prevent the addition of unused features

0616 GitHub stars

Use Cases

01Processing complex feedback on Pull Requests from multiple team members

02Maintaining architectural integrity when receiving external security or performance audits

03Ensuring technical rigor during collaborative pair programming sessions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add lifangda/claude-plugins receiving-code-review

For use in Claude.ai and ChatGPT

Download Skill