Challenge product, design, and technical plans with clarification questions.
npx skills add https://github.com/luan/agents --skill grill-me-luan
Please help me install this Agent Skill. Skill: grill-me Source: https://github.com/luan/agents/tree/main/skills/grill-me Command: npx skills add https://github.com/luan/agents --skill grill-me-luan
This Skill prevents vague or risky plans from moving forward by forcing early clarification through relentless, assumption-checking questions.
Ask the AI to grill your plan and resolve each dependency one-by-one, using the codebase whenever possible, until you share a clear understanding.
High-intent search queries and answers about installing and using this skill.
To test assumptions in a product proposal, you challenge your plan with focused clarification questions and resolve dependencies branch-by-branch to reach shared understanding before implementation. This prevents building the wrong thing by forcing early verification.
The best way to review a technical plan and verify dependencies is through small-batch interrogation, walking down the design tree to resolve decisions one-by-one while prioritizing codebase inspection over speculation.
During a design critique, codebase inspection works by exploring the repository to answer design questions directly, prioritizing actual codebase exploration over speculation to uncover hidden assumptions and reduce risk.
Yes, you can use assumption testing for design walkthroughs by applying relentless plan interrogation to challenge every aspect of the design until the team reaches a shared, verified understanding.
A limitation of small-batch questioning for plan review is that it requires handling questions in concise, grouped batches to maintain momentum, meaning complex dependency mapping must be resolved incrementally rather than all at once.
lukeyliasInterview users one question at a time to clarify plans and designs.
kcrleeInterrogate plans and designs through structured branch-by-branch questioning.
zhengfranInterview users one question at a time to clarify plans and resolve decision branches.
coco-researchScans a codebase to generate focused assessment documents in .planning/codebase/.
