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

•

Collaboration Tools

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

About

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.

Key Features

Structured handoff to deep-thinking agents for final problem solving
Automated scoring system based on issue severity (High, Med, Low)
Integrated technical fact-checking to verify framework and library claims
Gamified 'promotion' incentive to maximize bug discovery and thoroughness
Dual-agent parallel processing with specialized arch and impl roles
1 GitHub stars

Use Cases

Reviewing mission-critical pull requests for hidden bugs and structural debt
Fact-checking technical assumptions and version compatibility in complex environments
Auditing large-scale architectural changes before they are implemented

About

Key Features

Structured handoff to deep-thinking agents for final problem solving
Automated scoring system based on issue severity (High, Med, Low)
Integrated technical fact-checking to verify framework and library claims
Gamified 'promotion' incentive to maximize bug discovery and thoroughness
Dual-agent parallel processing with specialized arch and impl roles
1 GitHub stars

Use Cases

Reviewing mission-critical pull requests for hidden bugs and structural debt
Fact-checking technical assumptions and version compatibility in complex environments
Auditing large-scale architectural changes before they are implemented