batch-grill-me

Interviews users in rounds of frontier questions to map a design decision tree.

3|1|Updated May 29, 2026
One-click install
npx skills add https://github.com/mambo-wang/CodingHub --skill batch-grill-me-mambo-wang
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: batch-grill-me
Source: https://github.com/mambo-wang/CodingHub/tree/main/.codebuddy/skills/batch-grill-me
Command: npx skills add https://github.com/mambo-wang/CodingHub --skill batch-grill-me-mambo-wang

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Design discussions often leave decisions silently assumed or half-explored, leading to misaligned plans discovered too late. This Skill structures the conversation as a design tree and relentlessly interviews the user until every branch is visited and a shared understanding is confirmed. ## Core Features & Use Cases - Design Tree Mapping: Models every decision as a node whose dependent sub-decisions branch off it, so nothing is silently assumed. - Frontier-Based Rounds: Asks all currently unblocked questions in one numbered batch with recommended answers, then recomputes the frontier after each round of user replies. - Fact Delegation: Dispatches sub-agents to look up environmental facts (filesystem, tools) instead of asking the user for anything the AI can find itself. - Use Case: Before starting a new feature, run this Skill to surface every open design decision — data model, API shape, error handling — round by round until the plan is fully specified and confirmed. ## Quick Start Ask the assistant to grill you in rounds about your project plan until every design decision is settled and confirmed.

Frequently Asked Questions about batch-grill-me

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

FAQPage Schema
How do I get an AI to ask me all the right design questions before coding?▼

Use a frontier-based interview approach where the AI maps your decisions as a tree and asks every currently unblocked question in numbered rounds. Each round of your answers unlocks the dependent questions for the next round until nothing remains assumed.

What is a design decision tree in AI-assisted planning?▼

A design decision tree models each choice as a node with dependent sub-decisions branching from it. The frontier is the set of questions whose prerequisites are already settled, so they can be asked immediately without guessing at unresolved answers.

Can the AI look up facts itself instead of asking me during planning?▼

Yes. When a question needs a fact from the environment such as filesystem contents or tool availability, the Skill dispatches a sub-agent to find it. Only genuine decisions are put to the user; lookups never block unrelated questions.

When does the interview process end?▼

The session ends when the frontier is empty, meaning every branch of the design tree has been visited and no decision is left silently assumed. The AI does not act on the plan until you explicitly confirm shared understanding.

What are the limitations of round-based design interviews?▼

The approach requires the user to answer each round before dependent questions can be asked, so it is slower than a single-pass plan. It is best suited to complex designs where missed assumptions are costly, not trivial tasks.