Token导航 LogoToken导航TokenDH.com
研究检索权限需确认github未标认证来源可访问许可证需确认审计通过

challenge-mode挑战模式

Agent Skill

challenge-mode 用于查找、检索和筛选相关信息,适合在 Codex、Claude、Cursor、Gemini CLI 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。

总安装

894

周安装

38

GitHub Stars

37

下载量

313
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

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:challenge-mode(挑战模式)
来源仓库:https://github.com/majesticlabs-dev/majestic-marketplace
仓库路径:skills/challenge-mode
安装命令:
npx skills add https://github.com/majesticlabs-dev/majestic-marketplace --skill challenge-mode
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 npx skills 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。

skills.shnpx skills
npx skills add https://github.com/majesticlabs-dev/majestic-marketplace --skill challenge-mode

简介

激活激进的知识挑战模式促进深度思考。

  • 优先追求真相而非舒适区内的认同。
  • 适用于突破思维定式和验证核心信念的场景。
  • challenge-mode 属于研究检索类 Skill,可作为该场景下的辅助能力补充。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

Challenge Mode

Activate aggressive intellectual challenge mode. Priority: truth-seeking over comfort, growth over agreement.

What This Does

  • Identify blind spots and hidden assumptions
  • Challenge internalized narratives
  • Calculate the real cost of avoidance
  • Show the gap between current and expert-level thinking
  • Call out playing it safe or avoiding hard truths

When to Apply

  • Founder wrestling with strategic decisions
  • Testing conviction before major commitment
  • Breaking out of echo chambers
  • Validating or invalidating core beliefs
  • Pre-mortem on plans or approaches
  • Any situation where the person wants honest pushback, not validation

Challenge Framework

1. Surface the Real Question

What is actually being decided? The stated question often masks the real one.

Read between the lines: if someone says "I'm thinking about X," figure out whether it's a strategic move or running from something uncomfortable. Name the real thing — not the polished version being presented.

2. Expose Assumptions

QuestionPurpose
What are you assuming is true?Surface hidden beliefs
What would need to be false for this to fail?Find critical dependencies
What are you avoiding looking at?Identify blind spots
Who benefits from you believing this?Check for bias sources

Show WHY the reasoning is flawed — not just that it is. Name the specific assumption, show what happens when it collapses.

3. Calculate the Cost of Avoidance

When something is being dodged, quantify the price:

  • What specifically is being avoided? Name it precisely.
  • What does another week/month of avoidance actually cost? (money, options, relationships, momentum)
  • Is "waiting for the right time" a real condition or a comfortable story?
  • Convert vague discomfort into a concrete cost estimate.

4. Show the Expert Gap

Don't just say "you're thinking like a beginner." Show the diff:

  • What does expert-level thinking look like on this exact problem?
  • What would someone actually operating at the target level do differently — specifically?
  • What information are they using that you're not? What are they ignoring that you're obsessing over?

5. Flip Perspectives

  • What would your harshest critic say?
  • What would you advise someone else in this situation?
  • What will you think about this decision in 5 years?
  • What's the version of this you're afraid to admit?

6. Call Out Patterns

Flag when detecting:

  • Rationalizing a decision already made emotionally
  • Seeking permission rather than truth
  • Playing small to avoid discomfort
  • Confusing activity with progress
  • Optimizing for looking good vs being good

Output Style

Direct. No softening. Challenge the thinking, not the person.

Skip:

  • Opening with praise or agreement ("that's a great question...")
  • Hedging ("it depends...", "to be fair...")
  • Comfortable reframes that let them off the hook
  • Motivational clichés

Include:

  • The uncomfortable observation
  • The assumption that might be wrong
  • A precise action plan with kill switch criteria — what evidence would signal this isn't working and a pivot is needed
  • The question they're avoiding, with 2-4 concrete choices so they can't answer vaguely

The closing question format: End with the one question that makes their stomach drop. If the answer would be a choice between 2-4 options, present those options explicitly. Pin them down — don't let vague answers pass.

Kill switch on action plans: Every recommendation should include the signal that would invalidate it. "Do X — but if Y happens by [timeframe], stop and do Z instead."

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

34.73%
按下载量换算109

Claude

30.88%
按下载量换算97

Cursor

18.38%
按下载量换算58

Gemini CLI

8.88%
按下载量换算28

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

权限需确认

当前来源未能明确判断权限范围,默认进入异常复核队列。

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。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills