creview-spec

Coordinates a five-agent adversarial review of software specifications to surface hidden issues.

65|2|Updated Mar 27, 2026
One-click install
npx skills add https://github.com/joshft/correctless --skill creview-spec
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: creview-spec
Source: https://github.com/joshft/correctless/tree/main/correctless/skills/creview-spec
Command: npx skills add https://github.com/joshft/correctless --skill creview-spec

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill coordinates a multi-agent adversarial review of a software specification to identify hidden assumptions, gaps, and edge cases that a single reviewer might miss.

Core Features & Use Cases

  • Spawns specialized agents (Red Team, Assumptions Auditor, Testability Auditor, Design Contract Checker, Upgrade Compatibility Auditor) to review a spec from different perspectives.
  • Tracks progress, surfaces findings in real time, and synthesizes results into actionable changes for the spec.
  • Use Case: For critical feature specs, run creview-spec after /cspec or /cmodel to validate robustness before implementation.

Quick Start

Invoke the creview-spec workflow with /creview-spec to start the five-agent adversarial review.

Frequently Asked Questions about creview-spec

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

FAQPage Schema
How do I run an adversarial architecture review on a software spec?

An adversarial spec review uses five specialized agents including a Red Team and an Assumptions Auditor to analyze a software spec from different perspectives, surfacing hidden gaps and edge cases that a single reviewer might miss.

When should I use multi-agent spec review for feature briefs?

Multi-agent spec review is designed for high-stakes feature briefs, architecture proposals, or change requests where early validation of robustness and risk identification matters before implementation begins.

Can I automate risk identification for change requests before implementation?

Yes, you can automate risk identification by running creview-spec after generating your change request, which coordinates multiple agents to synthesize actionable changes and surface findings in real time.

What perspectives does a multi-agent adversarial analysis cover?

A multi-agent adversarial analysis covers five perspectives: Red Team, Assumptions Auditor, Testability Auditor, Design Contract Checker, and Upgrade Compatibility Auditor, synthesizing their findings into actionable spec modifications.

Does multi-agent spec review require any external dependencies?

No external dependencies are required to run the multi-agent spec review; the workflow operates independently to coordinate the five agents and synthesize their findings with mandatory progress visibility.

What's the best way to validate testability and upgrade compatibility in a spec?

The best way to validate testability and upgrade compatibility is using creview-spec, which deploys dedicated auditors to check these specific design contracts and synthesize the results into actionable changes.