How does it handle conflicting results?

The skill includes a merge and score step that deduplicates findings and prioritizes them based on severity, ensuring you get a clear, actionable summary of all verified issues.

What is the Competitive Reviewer skill?

It is a specialized Claude Code skill that dispatches two competing AI agents—an architect and an implementation expert—to find issues in your code or technical plan before you proceed with implementation.

When is the best time to use this skill?

It is most effective before starting complex tasks like modifying existing architecture, making major technical decisions, or building new features from scratch.

Why does the competition aspect matter?

Using a competitive 'promotion' framework incentivizes the AI agents to be more exhaustive and critical, catching nuanced issues that a single-pass review might miss.

Competitive Reviewer

Name: Competitive Reviewer
Author: ANcpLua

byANcpLua

•

协作工具

Deploys dual AI reviewers in a competitive framework to identify architectural flaws and implementation bugs before finalizing technical decisions.

The Competitive Reviewer skill optimizes Claude's analytical process by instigating a 'survival of the fittest' review between two specialized agents: an architecture reviewer and an implementation reviewer. By framing the task as a competition where the agent finding more high-severity issues is 'promoted,' the skill drives extreme thoroughness in catching structural weaknesses and code-level bugs. It automates the parallel dispatching, deduplication, and fact-checking of results, providing a high-fidelity context foundation for complex coding tasks or architectural refactoring.

主要功能

01Structured handoff to deep-thinking agents for final problem solving

02Automated scoring system based on issue severity (High, Med, Low)

03Integrated technical fact-checking to verify framework and library claims

04Gamified 'promotion' incentive to maximize bug discovery and thoroughness

05Dual-agent parallel processing with specialized arch and impl roles

061 GitHub stars

使用场景

01Reviewing mission-critical pull requests for hidden bugs and structural debt

02Fact-checking technical assumptions and version compatibility in complex environments

03Auditing large-scale architectural changes before they are implemented

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add ancplua/ancplua-claude-plugins competitive-review

For use in Claude.ai and ChatGPT

Download Skill