Token导航 LogoToken导航

Agent Skill · 研究检索

stuck(stuck 文档)

stuck 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。

stuck 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证为 MIT,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 47,393。

总安装 734周安装 30GitHub Stars 23下载 238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

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:stuck(stuck 文档)
来源仓库:https://github.com/johnlindquist/claude
仓库路径:skills/stuck
安装命令:
npx skills add https://github.com/johnlindquist/claude --skill stuck
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。不同来源提供的安装方式可能略有差异;本站展示可直接复制的安装命令,安装前请核对来源页面。

skills.shnpx skills
npx skills add https://github.com/johnlindquist/claude --skill stuck

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。

stuck 用于查找、检索和筛选相关信息。适合在 Codex、Claude、Cursor、Gemini CLI 中根据关键词、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。

安装命令:npx skills add https://github.com/johnlindquist/claude --skill stuck

来源仓库内路径:skills/stuck

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 Codex、Claude、Cursor、Gemini CLI;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 npx skills add https://github.com/johnlindquist/claude --skill stuck;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

stuck 怎么安装?

在终端执行 npx skills add https://github.com/johnlindquist/claude --skill stuck。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

stuck 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。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。

stuck 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。

stuck 是免费的吗?

来源标注的许可证是 MIT。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。

stuck 用的人多吗?

公开数据是累计安装 734 次,近一周安装 30 次,下载 238 次,GitHub 23 星。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。

stuck 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

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 3 个目录同时收录,说明这个 Skill 流传比较广。
  • 许可证明确为 MIT,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。

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 Codex,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
stuck

stuck 文档 · 本页

73423只读MIT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

Stuck Helper

Systematic approaches for getting unstuck on hard problems.

When to Use

  • You've tried multiple approaches without success
  • The same error keeps appearing despite fixes
  • You don't understand why something isn't working
  • Debugging has hit a dead end

Quick Unstuck

gemini -m pro -o text -e "" "I'm stuck on this problem:

PROBLEM: [describe what you're trying to do]
TRIED: [what you've already attempted]
ERROR/SYMPTOM: [what's happening]
EXPECTED: [what should happen]

Help me get unstuck. Consider:
1. Am I solving the right problem?
2. What assumptions might be wrong?
3. What haven't I tried?
4. What would an expert check first?"

Structured Debugging

Rubber Duck Protocol

gemini -m pro -o text -e "" "I need to rubber duck debug this issue:

I'm trying to: [goal]
The code does: [behavior]
I expected: [expected behavior]
The relevant code is:

\`\`\`
[paste code]
\`\`\`

Walk through the code step by step with me, questioning each assumption."

Assumption Audit

gemini -m pro -o text -e "" "Audit my assumptions about this problem:

PROBLEM: [description]
MY ASSUMPTIONS:
1. [assumption 1]
2. [assumption 2]
3. [assumption 3]

For each assumption:
- Is it definitely true?
- How would I verify it?
- What if it's wrong?"

Fresh Perspective

gemini -m pro -o text -e "" "I've been stuck on this for a while. Give me a completely fresh approach:

PROBLEM: [description]
WHAT I'VE TRIED:
- [approach 1]
- [approach 2]
- [approach 3]

Suggest approaches from a different angle. What would someone with no context try?"

Common Stuck Scenarios

Async/Promise Issues

gemini -m pro -o text -e "" "Debug this async code issue:

CODE:
\`\`\`
[paste async code]
\`\`\`

SYMPTOM: [what's happening]

Check for:
- Missing await
- Race conditions
- Unhandled rejections
- Callback timing
- Promise chain issues"

State Management

gemini -m pro -o text -e "" "Debug this state management issue:

FRAMEWORK: [React/Vue/etc]
CODE:
\`\`\`
[paste code]
\`\`\`

SYMPTOM: State not updating / Stale state / Infinite loop

Check for:
- Mutation vs immutable update
- Dependency arrays
- Closure over stale values
- Re-render triggers"

Type Errors

gemini -m pro -o text -e "" "Help me understand this TypeScript error:

ERROR:
\`\`\`
[paste error]
\`\`\`

CODE:
\`\`\`
[paste code]
\`\`\`

Explain:
1. What the error means in plain English
2. Why TypeScript is complaining
3. How to fix it properly (not just any)"

Build/Config Issues

gemini -m pro -o text -e "" "Debug this build/config issue:

TOOL: [webpack/vite/esbuild/etc]
ERROR:
\`\`\`
[paste error]
\`\`\`

CONFIG:
\`\`\`
[paste config]
\`\`\`

Common causes and solutions for this type of error."

Escalation Ladder

When simple debugging fails, escalate systematically:

Level 1: Verify Environment

# Check versions
node --version
npm --version
# Check dependencies
npm ls
# Clean and reinstall
rm -rf node_modules package-lock.json && npm install

Level 2: Minimal Reproduction

# Create minimal test case
mkdir /tmp/test-issue
cd /tmp/test-issue
# Create smallest code that reproduces issue

Level 3: Git Bisect

# Find when it broke
git bisect start
git bisect bad HEAD
git bisect good <last-known-good-commit>
# Test each commit git suggests

Level 4: Search for Similar Issues

# GitHub issues
gh search issues "[error message]" --limit 10

# With specific repo
gh search issues "[error]" --repo package/repo

Level 5: Ask for Help

# Prepare a good question
gemini -m pro -o text -e "" "Help me write a good Stack Overflow question for:

PROBLEM: [description]
CODE: [minimal reproduction]
ERROR: [full error]
TRIED: [what I've attempted]

Format as a clear, answerable question."

Prevention Checklist

After getting unstuck, prevent recurrence:

gemini -m pro -o text -e "" "I just solved this issue:

PROBLEM: [what was wrong]
SOLUTION: [how I fixed it]

Suggest:
1. Test to prevent regression
2. Documentation to add
3. Linting/tooling to catch this earlier
4. Pattern to follow in future"

Best Practices

  1. State the problem clearly - Writing it out often reveals the issue
  2. List what you've tried - Prevents repeating failed approaches
  3. Question assumptions - Often the issue is a wrong assumption
  4. Simplify - Remove code until you find the minimal case
  5. Take breaks - Fresh eyes see what tired ones miss
  6. Ask for help early - Don't spend hours on a 5-minute question
  7. Document the solution - Future you will thank present you

平台分布

Claude Code

27.52%
按下载量换算65

OpenCode

23.28%
按下载量换算55

Antigravity

17.44%
按下载量换算42

Gemini CLI

13.38%
按下载量换算32

windsurf

7.34%
按下载量换算17

trae

3.6%
按下载量换算9

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

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。

继续浏览同类 Skills