critique

Spawns auditors to explore codebases and verify claims in plans or specs.

13|1|Updated Feb 14, 2026
One-click install
npx skills add https://github.com/davidabeyer/formaltask --skill critique-davidabeyer
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: critique
Source: https://github.com/davidabeyer/formaltask/tree/main/skills/critique
Command: npx skills add https://github.com/davidabeyer/formaltask --skill critique-davidabeyer

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill acts as an adversarial reviewer, rigorously examining plans, specifications, or task artifacts to identify flaws, inconsistencies, and potential issues before they impact development.

Core Features & Use Cases

  • Automated Auditing: Spawns specialized auditors to explore the codebase and verify claims made in plans or specs.
  • Conflict Resolution: Reconciles conflicting findings from multiple auditors and synthesizes a unified critique.
  • Use Case: Before committing to a new feature, run this Skill on the proposed plan to ensure all acceptance criteria are testable, dependencies are correctly identified, and potential blast radius issues are flagged.

Quick Start

Critique the project plan located in the 'my-feature' directory.

Frequently Asked Questions about critique

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

FAQPage Schema
How do I run an adversarial code review on a project plan before implementation?

Adversarial code review audits plans by spawning specialized auditors to explore the codebase, verify task artifacts, validate paths, and synthesize findings into an approval verdict before development begins.

What is automated spec verification and how does it handle conflicting findings?

Automated spec verification spawns multiple auditors to check acceptance criteria and dependencies, then reconciles conflicting findings to synthesize a unified critique and final approval verdict for the plan.

Can I use automated auditing to check task readiness and commit approved tasks to a database?

Yes, automated auditing verifies task readiness by validating paths and context symbols, resolving auditor conflicts, and committing ready tasks to a database upon determining a final approval verdict.

When do I need plan critique for software engineering tasks?

Plan critique is needed before committing to a new feature to ensure acceptance criteria are testable, dependencies are correctly identified, and potential blast radius issues are flagged in the codebase.

Does this codebase analysis approach work without external dependencies?

Yes, codebase analysis operates without external dependencies, running self-contained scripts and references to perform task verification, path validation, and context symbol checks for specs.

What is the best way to flag blast radius issues in a proposed feature specification?

To flag blast radius issues, run an adversarial review on the feature specification to rigorously examine plans, identify inconsistencies, and verify all claims against the actual codebase.