deep-interview

Agent Skills › jjongguet/oh-my-product › deep-interview

deep-interview

GitHub

在需求模糊时通过苏格拉底式提问澄清目标、约束和验收标准，输出结构化摘要以指导后续执行。

src/skills/deep-interview/SKILL.md jjongguet/oh-my-product

Trigger Scenarios

需求宽泛或定义不清存在多种解读可能权衡关系不明确用户要求进行需求澄清

Install

npx skills add jjongguet/oh-my-product --skill deep-interview -g -y

More Options

Non-standard path

npx skills add https://github.com/jjongguet/oh-my-product/tree/main/src/skills/deep-interview -g -y

Use without installing

npx skills use jjongguet/oh-my-product@deep-interview

指定 Agent (Claude Code)

npx skills add jjongguet/oh-my-product --skill deep-interview -a claude-code -g -y

安装 repo 全部 skill

npx skills add jjongguet/oh-my-product --all -g -y

预览 repo 内 skill

npx skills add jjongguet/oh-my-product --list

SKILL.md

Frontmatter

{
    "name": "deep-interview",
    "aliases": [
        "\/deep-interview",
        "deep interview",
        "interview",
        "requirements interview"
    ],
    "description": "Run a structured Socratic interview before planning or execution.",
    "primaryRole": "analyst"
}

Deep Interview clarifies goals, constraints, and success criteria before implementation. Use it when ambiguity is high and premature coding would create rework.

<Use_When>

The request is broad, vague, or underspecified
Multiple interpretations are possible
Tradeoffs (time, quality, risk, scope) are unclear
The user asks for an interview, discovery, or requirements clarification </Use_When>

<Do_Not_Use_When>

The task is a small focused fix with clear acceptance criteria
The user explicitly asks to skip discovery and proceed immediately </Do_Not_Use_When>

<Execution_Policy>

Ask one question at a time
Prefer concrete, decision-driving questions
Converge quickly: stop when requirements are actionable
Finish with an explicit contract summary and open risks </Execution_Policy>

1. Restate the objective and constraints from the prompt. 2. Ask targeted questions in this order: - Outcome and non-goals - Scope boundaries - Operational constraints (time, infra, compliance) - Acceptance criteria and evidence 3. After each answer, refine assumptions and ask the next highest-impact question. 4. Stop when you can produce a clear implementation-ready brief. 5. Output a final summary with: - Objective - In-scope / out-of-scope - Acceptance criteria - Risks / unknowns - Recommended next step (`plan`, `team`, or direct execution)

<Output_Format>

Interview Summary

Objective

Scope

In: ...
Out: ...

Acceptance Criteria

Open Risks / Unknowns

Recommended Next Action

Use omp skill plan (or another named next step) with the summary above. </Output_Format>

Task: {{ARGUMENTS}}

Version History

4f49e99 Current 2026-07-25 06:34

Metadata

Files: 0
Version: 4f49e99
Hash: 6e9f33da
Indexed: 2026-07-25 06:34

deep-interview

deep-interview

Trigger Scenarios

Install

SKILL.md

Interview Summary

Objective

Scope

Acceptance Criteria

Open Risks / Unknowns

Recommended Next Action

Version History

Same Skill Collection

Metadata

文库