Critique documents and plans to identify and prioritize weaknesses.
npx skills add https://github.com/dvy1987/agent-loom --skill adversarial-hat
Please help me install this Agent Skill. Skill: adversarial-hat Source: https://github.com/dvy1987/agent-loom/tree/main/.agents/skills/adversarial-hat Command: npx skills add https://github.com/dvy1987/agent-loom --skill adversarial-hat
It identifies unsupported claims, hidden assumptions, and fragile solutions in documents and plans so teams avoid committing to flawed strategies or designs.
Run an adversarial review on the attached PRD and return critical, significant, and minor findings with evidence and concrete resolution conditions.
High-intent search queries and answers about installing and using this skill.
To run an adversarial review, provide the document text, stated claims, and desired stakes level. The process performs a three-phase red team critique to identify unsupported claims and hidden assumptions, returning severity-tagged findings with evidence and explicit resolution conditions.
Red team critique is a structured adversarial analysis used to expose and fix weaknesses before commitment. You need it for product-soul validation, design docs, or implementation plans to verify claims, stress-test solutions, and avoid committing to fragile strategies.
Yes, you can use assumption-testing to prioritize weaknesses by supplying the implementation plan text and desired stakes level. The critique outputs severity-tagged findings—Critical, Significant, or Minor—based on evidence to help prioritize plan risks.
The adversarial critique process does not specify required document formats but requires the document text, stated claims or metrics, and desired stakes level as input to produce actionable findings and explicit resolution conditions.
The best way to stress-test failure modes is applying the Challenge phase of an adversarial critique. By providing the brainstorming output and claims, the process broadens problem framing and returns severity-tagged findings with concrete resolution conditions.
dvy1987Generate a Product Soul document across five strategic lenses to docs/product-soul.md.
ClyptAIGenerate a GitHub issue-ready PRD from conversation and codebase context.
StepowskiEricMap and invert failure modes into guardrails and a failure-map.md artifact.
BellaBeValidate assumptions and design experiments to de-risk product decisions.
