Token导航 LogoToken导航

Agent Skill · 研究检索

devils-advocate-suite(魔鬼代言人套房)

devils-advocate-suite 用于查找、检索和筛选相关信息,适合在 OpenClaw 中需要根据关键词、任务场景或来源线索快速定位候选结果时使用。可结合来源仓库、安装命令和原始 README 继续核验具体用法。安装前建议确认权限范围、维护状态,以及是否会触发联网、命令执行或文件读写。

devils-advocate-suite 是研究检索类 Agent Skill,可用宿主为 OpenClaw,通过ClawHub获取,权限为只读;许可证为 MIT-0,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 24,544。

总安装 3,990周安装 163GitHub Stars 公开资料未说明下载 1,278核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT-0
版本
1.0.0

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:devils-advocate-suite(魔鬼代言人套房)
来源仓库:https://github.com/midnightstudioai/devils-advocate-suite
安装命令:
openclaw skills install devils-advocate-suite
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 OpenClaw 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

ClawHubOpenClaw
openclaw skills install devils-advocate-suite

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

执行对抗性思维训练,对高风险决策进行系统性压力测试。适合在 OpenClaw 中需要打破群体思维、识别潜在风险或挑战假设时使用。提供多角度质疑框架与应急预案推演能力。安装命令:openclaw skills install devils-advocate-suite;无特殊资源消耗。

输出结论需结合正向视角综合评估,不可单独依赖。

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 OpenClaw;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 openclaw skills install devils-advocate-suite;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

devils-advocate-suite 怎么安装?

在终端执行 openclaw skills install devils-advocate-suite。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

devils-advocate-suite 支持哪些 AI 工具?

公开资料标注支持 OpenClaw。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

devils-advocate-suite 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

devils-advocate-suite 是免费的吗?

来源标注的许可证是 MIT-0。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

devils-advocate-suite 用的人多吗?

公开数据是累计安装 3,990 次,近一周安装 163 次,下载 1,278 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

devils-advocate-suite 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 专注 OpenClaw 一种宿主,触发方式和使用习惯不用来回切换。
  • 许可证明确为 MIT-0,商用和分发边界好判断。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 OpenClaw,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 需要许可证清楚、并且有第三方扫描结果的团队项目。

不太适合

  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
devils-advocate-suite

魔鬼代言人套房 · 本页

3,990—只读MIT-0通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

name
devils-advocate-suite
description
Execute structured adversarial thinking to stress-test high-stakes decisions, identify systemic risks, and prevent groupthink. Use when the user asks for a strategy review, risk assessment, pre-mortem, Amazon-style PR/FAQ review, or needs to challenge assumptions regarding a project or business model. Make sure to use this skill whenever a decision has dire consequences or shows signs of false consensus.
license
MIT

Devil's Advocate Suite: Operationalizing the Institutionalized Adversary

Work Philosophy

This skill is grounded in Adversarial Epistemology. It moves beyond casual contrarianism to provide a structured "override operation," forcing a "decoupling" of prior beliefs from the evaluation of evidence. Based on the research of Charlan Nemeth, this skill prioritizes authentic dissent over mere role-playing where possible, as authentic dissent is significantly more effective at stimulating divergent thinking and uncovering correct solutions. When role-playing is necessary, it employs specific protocols (Pre-Mortem, PR/FAQ, TRIZ) to prevent the "cognitive bolstering" of original views that often plagues assigned Devil's Advocates.

Operational Workflow

Phase 1: The Pre-Mortem (Prospective Hindsight)

Use this at the start of a project to increase the ability to identify future failure points by 30%.

  1. Scene Setting: State: "Imagine it is one year in the future. The project has failed spectacularly. It is an embarrassing and devastating disaster."
  2. Independent Generation: List all possible reasons for this failure silently for 2–5 minutes.
  3. Aggregation: Use python scripts/premortem_aggregator.py to organize these into a Risk Matrix (Likelihood vs. Impact).
  4. Mitigation: Develop specific counter-measures for high-impact/high-likelihood threats.

Phase 2: Narrative Stress-Testing (Amazon Protocol)

Used for strategic alignment and ruthless prioritization.

  1. The PR/FAQ: Draft a one-page Press Release (PR) for the "successful" launch and an Internal FAQ addressing the most brutal questions about TAM, P&L, and dependencies.
  2. Silent Start: Spend 20 minutes reading the narrative in silence before any discussion. This prevents "charismatic leader" bias.
  3. The "So What" Critique: Every technical detail must answer for its business value; every strategic goal must specify concrete actions.

Phase 3: Dialectical Inquiry (DI)

Use for complex, ill-structured problems where a "Third Path" is needed.

  1. Subgroup Split: Divide into "Proposal Team" and "Counter-Proposal Team."
  2. Assumption Breakdown: The Counter-Proposal team must invent plausible counter-assumptions to every core claim of the original plan.
  3. Synthesis: Search for a new, superior alternative that incorporates the best components of both worldviews.

Phase 4: TRIZ Contradiction Resolution

Transform the Devil's Advocate from a "naysayer" into a "catalyst" for innovation.

  1. Identify Contradictions: Use python scripts/triz_solver.py when an advocate identifies that improving one parameter (e.g., speed) worsens another (e.g., cost).
  2. Apply Principles: Use principles like "Segmentation," "Taking Out," or "Intermediary" to solve the technical or business conflict.

Specific Commands

  • analyze --method premortem: Executes the 7-step failure simulation.
  • analyze --method prfaq: Structures a strategy review around an Amazon-style narrative.
  • solve --contradiction [Param A] vs [Param B]: Invokes the TRIZ solver.

References

Consult references/methodology_summary.md for full theoretical grounding on Nemeth dissent research, Klein pre-mortem, Amazon Working Backwards, Janis groupthink prevention, and TRIZ contradiction resolution.

平台分布

OpenClaw

95.82%
按下载量换算1,225

安全审计

VirusTotal

通过

ClawScan

通过

Static analysis

通过

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills