interrogate-me

Resolves load-bearing decisions in plans through sequential one-at-a-time questioning.

Updated Jul 24, 2026
One-click install
npx skills add https://github.com/urban233/ntn-ingest --skill interrogate-me-urban233
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: interrogate-me
Source: https://github.com/urban233/ntn-ingest/tree/main/.agents/skills/interrogate-me
Command: npx skills add https://github.com/urban233/ntn-ingest --skill interrogate-me-urban233

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Plans, ideas, and implementation approaches often contain unresolved decisions that surface too late. This Skill stress-tests a plan by interrogating it one question at a time, each with a recommended answer, until every load-bearing decision is resolved and a shared understanding is confirmed. ## Core Features & Use Cases - Sequential decision interrogation: Asks the single most important open question at a time, always paired with a recommended answer and reasoning, so the user only confirms, rejects, or modifies. - Decision tree traversal: Follows each answer's branch to its conclusion, surfaces dependencies between decisions, and flags conflicts with earlier choices. - Confirmed decision summary: Produces a numbered list of every decision with rationale, marks unresolved items as deferred, and waits for explicit user confirmation before any next step. - Use Case: Before drafting a feature spec, invoke this Skill to walk through the issue's open implementation questions, resolve each one in dependency order, and produce a confirmed decision record for the spec author. ## Quick Start Ask the assistant to interrogate your plan or decision, for example by saying "interrogate me on my plan to migrate the database schema".

Frequently Asked Questions about interrogate-me

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

FAQPage Schema
How do I stress-test a plan before implementing it?▼

Invoke the Skill with your plan stated in one or two sentences. It identifies the most foundational decision first, asks one question at a time with a recommended answer, and follows each branch until all decisions are resolved and confirmed.

How does the interrogate-me skill work with feature-spec-creator?▼

When invoked compositionally, it loads the issue's open questions for implementation as seed questions instead of restarting elicitation. It orders them by dependency, resolves each with recommendations, and outputs a confirmed decision summary for spec drafting.

Does the skill make decisions on my behalf?▼

No. It always presents a recommended answer with reasoning, but the user confirms, rejects, or modifies each decision. It never acts on decisions and waits for explicit confirmation of the final summary before proceeding.

What happens when a decision cannot be resolved during interrogation?▼

The question is explicitly flagged as deferred with a note to resolve it before the spec is finalized or that part of the implementation begins. Uncertainty is surfaced rather than pushed through.

When should I not use one-question-at-a-time decision interrogation?▼

Avoid it for trivial decisions with no downstream dependencies or when facts can simply be looked up in the codebase. The Skill itself checks facts first and reserves questions strictly for genuine decisions.