跳到主要内容
知仓学习社ZHICANG

devils-advocate

Challenge research design decisions, assumptions, and methodology choices with specific critical questions. Helps strengthen the paper before submis…

不碰外部(只输出文字)无严重或高危命中brycewang-stanford/Auto-Empirical-Research-Skills

它会碰到什么

扫了多少1 个文本文件,2 KB
它会碰到什么不碰外部(只输出文字)
命中总数0 处
命中统计严重 0 · 高 0 · 中 0 · 低 0

这一栏是扫描器报的事实,不是结论。命中多不等于有毒(安全工具、规则库、示例脚本本来就会包含危险写法),命中少也不等于干净。它和你手上的凭据、文件、网络有什么关系,需要你自己看。

技能内容

Devil's Advocate

Critically examine a research design, methodology, or assumption with 5-7 specific challenges.

Steps

  1. Understand the target: Read $ARGUMENTS and any relevant project files (CLAUDE.md, manuscript, estimation code).
  1. Generate 5-7 challenges from these categories:
  • Identification: Could the causal claim be undermined by [specific threat]?
  • Specification: What if the functional form is wrong? Alternative specifications?
  • Data: Are there sample selection issues? Measurement error concerns?
  • Methodology: Could a different estimator give different results? Why this method?
  • Assumptions: Which assumptions are most vulnerable? What if they're violated?
  • Robustness: What robustness checks are missing?
  • External validity: Does this generalize beyond the sample?
  1. For each challenge, provide:
  • The question
  • Why it matters
  • Suggested resolution or robustness check
  • Severity (Critical / Major / Minor)
  1. Summary verdict: Strengths, critical changes needed, and suggested improvements.

Notes

  • Be constructive, not destructive.
  • Think like a skeptical referee at a top journal.
  • Focus on the most impactful challenges first.

想直接用这个技能?

本站把开放许可(MIT / Apache 等)的技能按仓库打包整理到网盘,点一下转存到你自己的网盘,不用一个个从 GitHub 拉。许可未声明的技能只给原始仓库链接,不打包。

同名技能的其他版本

有 3 个不同仓库或目录里都有叫 devils-advocate 的技能。它们内容并不相同,别混用: