adversarial-review

Dispatch a cold subagent to contest assumptions in technical plans.

Updated Apr 17, 2026
One-click install
npx skills add https://github.com/xinye1/xl-skills --skill adversarial-review-xinye1
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: adversarial-review
Source: https://github.com/xinye1/xl-skills/tree/main/skills/adversarial-review
Command: npx skills add https://github.com/xinye1/xl-skills --skill adversarial-review-xinye1

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill overcomes the commitment bias inherent in self-authored plans and specifications by providing an objective, independent, and adversarial review process.

Core Features & Use Cases

  • Cold Reviewer Dispatch: Deploys a high-capability subagent to contest assumptions and identify flaws in design docs, implementation plans, and ADRs.
  • Structured Triage: Forces a rigorous, evidence-based triage process where findings are adjudicated against the artifact and codebase.
  • Use Case: Before starting a complex multi-phase migration, use this skill to have an independent agent stress-test your implementation plan, ensuring edge cases and architectural risks are identified before any code is written.

Quick Start

Trigger an adversarial review of the current implementation plan located at docs/plans/migration.md to identify potential architectural flaws.

Frequently Asked Questions about adversarial-review

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

FAQPage Schema
How do I stress-test a technical implementation plan for architectural flaws?

To stress-test an implementation plan, you can trigger an adversarial review that dispatches a cold subagent to contest assumptions and identify architectural risks before any code is written.

What is an adversarial review of a design specification?

An adversarial review is an independent assessment process that overcomes authoring bias by having a cold subagent rigorously scrutinize design specifications and verify claims against repository context.

When do I need to run an adversarial review on my architecture decisions?

You need an adversarial review before starting complex multi-phase migrations or committing to architectural decisions, ensuring edge cases and failure modes are identified early in the planning phase.

Can I use this adversarial review process on existing ADRs?

Yes, the adversarial review process supports design docs, implementation plans, and ADRs by deploying a high-capability subagent to triage findings and adjudicate claims against the codebase.

Do I need high-capability models to perform rigorous specification review?

Yes, rigorous specification review requires high-capability models to ensure the dispatched subagent can thoroughly scrutinize architectural decisions, contest assumptions, and verify complex failure modes.

What are the limitations of automated adversarial plan review?

Automated adversarial plan review is limited by its dependence on repository context for evidence-based triage, meaning it cannot effectively contest assumptions or verify claims for undocumented external systems.