plan-critic

Review implementation plans for security gaps, missing requirements, and resilience issues.

Updated Jul 3, 2026

Search Tags:#architecture #requirements #security analysis #ai systems #code planning #plan review

Ready to run

npx skills add https://github.com/Reidond/Banshee --skill plan-critic-reidond

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: plan-critic
Source: https://github.com/Reidond/Banshee/tree/main/.claude/skills/plan-critic
Command: npx skills add https://github.com/Reidond/Banshee --skill plan-critic-reidond

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill prevents flawed implementation plans from reaching users by performing adversarial reviews that identify security gaps, missing requirements, resilience issues, and unnecessary complexity before execution.

Core Features & Use Cases

Implementation Plan Review: Critically evaluates spec, design, and task documents against completeness, security, stability, and impact criteria.
AI System Risk Analysis: Detects AI-specific failure modes, integration risks, missing contracts, and unsafe assumptions in AI-powered workflows.
Use Case: Use this Skill before presenting a software delivery plan to uncover blockers, challenge architectural decisions, and improve acceptance criteria.

Quick Start

Ask the plan-critic skill to review the implementation plan for security, completeness, resilience, and architectural risks before sharing it with stakeholders.

Frequently Asked Questions about plan-critic

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

FAQPage Schema

How do I review an implementation plan for security gaps and missing requirements?▼

Reviewing an implementation plan for security gaps requires an adversarial evaluation against structured criteria covering completeness, stability, and impact analysis to identify flaws before execution. This process challenges architectural decisions and detects missing requirements.

What is the best way to catch architectural flaws before software implementation starts?▼

Catching architectural flaws before implementation involves critically evaluating design documents and task plans to identify overengineering, resilience issues, and unnecessary complexity. This ensures unsafe assumptions are uncovered prior to approval.

How do I validate AI system workflows for integration risks and unsafe assumptions?▼

Validating AI system workflows for integration risks requires analyzing AI-specific failure modes and missing contracts within the design. This detects unsafe assumptions and ensures rigorous validation of AI-powered workflows before execution.

Can I use structured criteria to detect overengineering in task plans and specifications?▼

Detecting overengineering in task plans uses structured review criteria to evaluate architectural decisions and eliminate unnecessary complexity. This ensures the design maintains stability without adding excessive components before approval.

Does plan review work for AI-integrated system workflows requiring rigorous validation?▼

Plan review works for AI-integrated system workflows by detecting AI-specific failure modes and integration risks during the evaluation. It validates missing contracts and unsafe assumptions to ensure workflow stability before implementation.

When should I not use an adversarial review for software architecture documents?▼

Adversarial review for software architecture documents should be avoided if the plan lacks sufficient detail for impact analysis and completeness evaluation. Without structured criteria, the review cannot effectively identify resilience issues or missing requirements.