adversarial-review

Spawn opposite-model reviewers to critique code diffs and synthesize verdicts.

9.0k|629|Updated Apr 16, 2025
One-click install
npx skills add https://github.com/mengxi-ream/read-frog --skill adversarial-review-mengxi-ream
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: adversarial-review
Source: https://github.com/mengxi-ream/read-frog/tree/main/.agents/skills/adversarial-review
Command: npx skills add https://github.com/mengxi-ream/read-frog --skill adversarial-review-mengxi-ream

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Adversarial code review uses opposite-model reviewers to challenge work and surface hidden issues before shipping.

Core Features & Use Cases

  • Spawn 1–3 reviewers on the opposite model to critique changes across distinct lenses.
  • Generate a synthesized verdict that guides decision-making without altering the original code.
  • Reference brain/principles and leverages reviewer-lenses to uncover edge cases and failures in diffs.

Quick Start

Provide a diffs bundle to trigger the opposite-model adversarial review.

Frequently Asked Questions about adversarial-review

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What is adversarial code review and how does it surface hidden flaws in diffs?

Adversarial code review spawns 1–3 reviewers on the opposite model to critique software changes across distinct lenses. It produces a synthesized verdict that guides decision-making and reveals hidden flaws before shipping.

How do I run an adversarial review on a large commit using the opposite model?

To run an adversarial review, provide a diffs bundle to trigger the opposite model's CLI to spawn reviewers. These reviewers adhere to defined lenses and brain principles to generate a synthesized verdict.

Can I use adversarial review for planning sessions or is it only for code diffs?

Adversarial review is applicable to both code diffs and planning sessions. By leveraging reviewer-lenses, the opposite-model reviewers can challenge work and uncover edge cases across various stages of development.

Does adversarial code review alter the original code in my large commits?

Adversarial code review does not alter the original code. It generates a synthesized verdict to guide decision-making, ensuring the author's work remains unchanged while hidden issues are surfaced.

What are the limitations of using opposite-model reviewers for code critique?

Using opposite-model reviewers requires the opposite model's CLI to spawn reviewers, meaning it depends on external CLI availability. It critiques changes without modifying code, leaving resolution to the author.