human-verify

Name: human-verify
Availability: InStock
Author: InerNoro

Review code changes via adversarial debate, reverse verification, and boundary testing.

2|1|Updated Dec 10, 2025

Search Tags:#verification #code-review #validation-report #boundary-testing #adversarial-review #scenario-simulation

Ready to run

npx skills add https://github.com/inernoro/prd_agent --skill human-verify

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: human-verify
Source: https://github.com/inernoro/prd_agent/tree/main/.claude/skills/human-verify
Command: npx skills add https://github.com/inernoro/prd_agent --skill human-verify

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Multi-perspective code verification that simulates human review. Combines adversarial debate (Devil's Advocate), reverse verification, boundary testing, and user scenario simulation to catch issues that compilers and linters miss. Trigger words: "/verify", "验证一下", "帮我检查".

Core Features & Use Cases

Adversarial debate orchestration to challenge proposed changes and surface hidden assumptions.
Reverse verification, boundary testing, and user-scenario simulation to validate functionality beyond surface correctness.
Structured outputs including validation reports and guidance for fixes to help teams ship safer code.

Quick Start

Trigger the verify flow on your latest change to generate a cross-perspective validation report.

Frequently Asked Questions about human-verify

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

FAQPage Schema

How do I catch edge cases that compilers and linters miss during code review?▼

To catch edge cases missed by compilers, perform multi-perspective code review using adversarial debate, reverse verification, boundary testing, and scenario simulation. This simulates human verification to detect hidden assumptions and surface issues before merging API endpoints, UI components, or logic refactors.

What is adversarial debate in code verification and how does it work?▼

Adversarial debate in code verification is a mechanism that challenges proposed changes to surface hidden assumptions. It orchestrates a Devil's Advocate perspective to simulate human review, identifying flaws in logic refactors and data processing that standard linters cannot detect.

How do I generate a validation report for my code changes before merging?▼

Generate a validation report by triggering the verify flow on your latest changes. The process performs boundary testing and reverse verification, outputting a structured validation report that includes identified issues and guidance for fixes to help teams ship safer code.

Does multi-perspective code review work for both API endpoints and UI components?▼

Yes, multi-perspective code review works for API endpoints, UI components, data processing, and logic refactors. It validates functionality beyond surface correctness by simulating user scenarios and boundary conditions, ensuring thorough verification whenever required before merging.

What is the best way to validate code refactoring beyond surface correctness?▼

The best way to validate code refactoring is using reverse verification and user-scenario simulation. By challenging the changes through adversarial review and boundary testing, you can validate functionality beyond surface correctness and generate a structured validation report.

When do I need scenario simulation for code verification?▼

You need scenario simulation for code verification whenever thorough validation is required before merging, especially across API endpoints, UI components, or data processing. It validates functionality beyond surface correctness by testing boundary conditions and user scenarios.