Runs a relentless interview session to sharpen a plan or design.
npx skills add https://github.com/nseng-ai/ns --skill grill-me-nseng-ai
Please help me install this Agent Skill. Skill: grill-me Source: https://github.com/nseng-ai/ns/tree/main/.agents/skills/grill-me Command: npx skills add https://github.com/nseng-ai/ns --skill grill-me-nseng-ai
/grilling session that probes a plan or design with persistent, pointed questions. - Assumption Surfacing: Forces the author to articulate and defend decisions that would otherwise remain implicit. - Use Case: Before starting a complex refactor, invoke the skill to have your design document interrogated question by question, revealing gaps in your reasoning while changes are still cheap to make. ## Quick Start Ask the agent to run a grilling session on your current plan or design document.High-intent search queries and answers about installing and using this skill.
Invoke a grilling session on your plan or design and answer the interview questions it poses. The relentless questioning format exposes hidden assumptions, vague reasoning, and unexamined tradeoffs before you commit to implementation.
A grilling session is a structured interview where the agent relentlessly questions your plan or design. Instead of passively reviewing the document, it probes decisions until your reasoning is explicit and defensible.
No external tools or dependencies are required. The skill consists only of conversational instructions in its SKILL.md file, with no scripts, references, or assets to install or configure.
Interview-style review is less useful when a design is already finalized and under time pressure, or when the problem is purely mechanical. It adds the most value early, while decisions can still be changed cheaply.
mattpocockInterviews a plan or design while generating ADRs and glossary documentation.
kashman001Interrogates a plan or design through relentless questioning to expose weaknesses.
DanhvbProvide ERP domain knowledge for requirements analysis across core modules.
LauughterSubject non-trivial outputs to fresh-context adversarial scrutiny through a five-step doubt cycle.