Thinking Methods

MCP办公协作

提供方 未标注

办公协作 stdio · 本地进程 会话登录 社区 / 第三方 ClaudeCursorVS CodeWindsurf
查看接入方式

面向AI代理的结构化思维技能:头脑风暴、决策支持、系统思维与策略。专为Claude Code和Cursor打造。

一句话选型结论

按资料字段整理

Thinking Methods 属于办公协作类 MCP 服务器,stdio · 本地进程,会话登录,可直接接入 Claude、Cursor、VS Code 等,有可直接执行的安装命令。接入前先确认权限范围;涉及写入、删除或付款的操作,建议先在测试环境验证。

分类办公协作 传输方式stdio · 本地进程 鉴权会话登录 部署方式本地部署
GitHub 星标 未标注 仓库热度,无公开仓库时显示未标注
工具数 未标注 暴露给模型的工具数量
资源 / 提示词 0 / 0 资源与提示词模板数量
维护方 社区 / 第三方 核验等级 仅资料核对

工具与能力

办公协作
数量来自来源目录的标注,具体工具清单以官方文档为准
工具数量未标注暴露给模型可调用的工具个数
资源数量0可读取的资源条目
提示词数量0内置提示词模板数量

能力标签

支持客户端

适用场景

生产力

安装与配置

stdio · 本地进程 会话登录
客户端在本机拉起子进程,数据不出本机
npx skills add ysskrishna/ai-agent-skills

这条命令只是入口,具体启动参数请按官方文档补齐,不要直接照搬到别的服务器。

  1. 1
    准备运行环境

    stdio 服务在本地起进程,需要先装好 Node.js 等运行时,再按安装命令拉取依赖。

  2. 2
    安装服务

    复制上面的安装命令执行;命令里带占位符的位置,换成自己的密钥或路径。

  3. 3
    写入客户端配置

    在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,让新配置生效。

  4. 4
    验证是否可用

    连接成功后随便调一个只读工具,确认能返回结果,再接入正式流程。

典型使用场景

3 类
按分类与标签整理的常见用法,具体以官方文档为准
日程与消息

查日程、发消息、约会议。

表单与审批

填表单、走审批流。

团队协作

同步任务、汇总进展。

选择建议

选之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 有现成的安装命令,不用自己找入口,照着跑就能起服务。
  • 已适配 Claude、Cursor、VS Code,换客户端不用重新研究配置。

先想清楚的地方

  • stdio 方式要求本地装好运行时(如 Node.js),换机器要重新装一遍。
  • 使用前要准备好会话登录对应的凭据,并按最小权限配置。
  • 资料没标注工具数量,实际能力要装好后确认。
  • 版本更新与行为细节以官方仓库为准。

更适合

  • 需要在办公协作场景里让模型直接调用外部能力的人。
  • 已经在用 Claude、Cursor 的用户,接入路径最短。
  • 想先快速试一下、能接受在本地或自有环境跑服务的人。

不太适合

  • 完全不想在本地装环境的人。
  • 不方便管理密钥或第三方授权的人。
  • 需要厂商级 SLA 和技术支持的生产环境。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
MCP 服务器工具数传输方式部署方式安装命令最后核验
Thinking Methods当前 未标注 stdio · 本地进程 本地部署 有 2026-11-20
CowAgent 未标注 stdio · 本地进程 本地部署 有 2026-04-11
Karakeep MCP 未标注 未标注 未标注 无 2026-05-17
agentscope 未标注 stdio · 本地进程 本地部署 有 2026-04-11
note-gen 未标注 stdio · 本地进程 本地部署 有 2026-04-11
Everywhere 未标注 stdio · 本地进程 本地部署 有 2026-04-12

服务器档案

办公协作
字段来自项目仓库与公开索引,中文说明由本站标注
提供方未标注上游标注的维护方 未标注
分类办公协作本站按上游分类展示
连接方式stdio · 本地进程客户端在本机拉起子进程,数据不出本机
鉴权方式会话登录远程服务建议用最小权限凭据
维护方社区 / 第三方核验等级 仅资料核对
包名 / 运行时未标注 · 未标注本地安装时用到的包标识与运行时
部署类型本地部署上游给出的部署形态
最近核验2026-11-20上游最后核验时间

详细介绍

15.3 KB
来自项目仓库与上游目录的原文,未做删改

Thinking Methods

A curated collection of cognitive workflows designed to upgrade your AI agents from simple code generators into strong collaborators for decision support, brainstorming, and structured thinking. Compatible with Claude Code, Codex, Gemini CLI, Cursor, Antigravity, GitHub Copilot CLI, Factory Droid, Qwen Code, Grok Build, OpenCode, Pi, Kimi Code, Hermes Agent, Devin, Muse, Windsurf, OpenClaw, and any tool that supports the same specification. See Installation.

Overview

Most coding assistants default to quick answers. This repository packages structured thinking methods as modular SKILL.md skills from the Agent Skills open standard, so an agent can run a method on a real decision instead of improvising one.

  • Describe the problem, not the method. Each skill's description lists the situations and phrases it fits ("what could go wrong?", "A or B?", "why does this keep happening?") and when to skip it. You do not need to name the method.
  • Not sure which one? Start with thinking-method-selector. It picks a method for your situation and runs it.
  • Each skill is standalone. Install one or all. Every skill has a light path for small questions and a worked example in its references/ folder.

Installation

Fastest path, works in 70+ agents:```bash
npx skills add ysskrishna/ai-agent-skills


Native installs for specific agents:

Claude Code```text
/plugin marketplace add ysskrishna/ai-agent-skills

# All skills in one plugin
/plugin install ai-agent-skills@ai-agent-skills

Or install single skills as separate plugins:```text
/plugin install thinking-method-selector@ai-agent-skills
/plugin install five-whys@ai-agent-skills
/plugin install analytical-thinking@ai-agent-skills
/plugin install systems-thinking@ai-agent-skills
/plugin install first-principles-thinking@ai-agent-skills
/plugin install design-thinking@ai-agent-skills
/plugin install swot-analysis@ai-agent-skills
/plugin install tradeoff-analysis@ai-agent-skills
/plugin install prioritization@ai-agent-skills
/plugin install strategic-thinking@ai-agent-skills
/plugin install fermi-estimation@ai-agent-skills
/plugin install critical-thinking@ai-agent-skills
/plugin install pre-mortem@ai-agent-skills
/plugin install ethical-thinking@ai-agent-skills
/plugin install creative-thinking@ai-agent-skills
/plugin install lateral-thinking@ai-agent-skills
/plugin install six-thinking-hats@ai-agent-skills


Update with `/plugin marketplace update ai-agent-skills`. Remove with `/plugin uninstall ai-agent-skills@ai-agent-skills`. Check with `/plugin`.

If the marketplace add fails with `Permission denied (publickey)`, run `git config --global url."https://github.com/".insteadOf git@github.com:` once. More in docs/claude-code-setup.md.

Codex CLI and App```bash
codex plugin marketplace add ysskrishna/ai-agent-skills
codex plugin add ai-agent-skills@ai-agent-skills

Start a new session, then describe your problem or call a skill with @. Update with codex plugin marketplace upgrade ai-agent-skills. Remove with codex plugin remove ai-agent-skills@ai-agent-skills. Needs Codex CLI v0.122 or later. More in docs/codex-setup.md.

Gemini CLI```bash

Extension: all skills in one install

gemini extensions install https://github.com/ysskrishna/ai-agent-skills

Or only the skills, without the extension wrapper

gemini skills install https://github.com/ysskrishna/ai-agent-skills.git --path skills


Check with `gemini skills list`. Update with `gemini extensions update ai-agent-skills`. Remove with `gemini extensions uninstall ai-agent-skills`. The extension install uses the latest GitHub Release. More in docs/gemini-cli-setup.md.

Cursor

In Cursor, open **Customize, Plugins, Add, From GitHub Repository** and enter `github.com/ysskrishna/ai-agent-skills`. In Agent chat you can also run `/add-plugin`.

Without the plugin route, install the skills into Cursor's skills folder:```bash
npx skills add ysskrishna/ai-agent-skills --agent cursor

Check under Customize, Skills. More in docs/cursor-setup.md.

Antigravity```bash
agy plugin install https://github.com/ysskrishna/ai-agent-skills


Reinstall with the same command to update. Remove with `agy plugin uninstall ai-agent-skills`. Skills are namespaced as `ai-agent-skills:`. More in docs/antigravity-setup.md.

GitHub Copilot CLI```bash
copilot plugin marketplace add ysskrishna/ai-agent-skills
copilot plugin install ai-agent-skills@ai-agent-skills

Check with copilot skill list. Update with copilot plugin update ai-agent-skills. Remove with copilot plugin uninstall ai-agent-skills. You can also install through GitHub CLI (see GitHub CLI). More in docs/copilot-cli-setup.md.

Factory Droid```bash
droid plugin marketplace add ysskrishna/ai-agent-skills
droid plugin install ai-agent-skills@ai-agent-skills --scope user


Install the `ai-agent-skills` bundle. The single-skill plugins load no skills on Droid. Update with `droid plugin update ai-agent-skills@ai-agent-skills`. Remove with `droid plugin uninstall ai-agent-skills@ai-agent-skills`. More in docs/droid-setup.md.

Qwen Code```bash
qwen extensions install ysskrishna/ai-agent-skills:ai-agent-skills

Skills appear as ai-agent-skills:. The install uses the latest GitHub Release. Update with qwen extensions update ai-agent-skills. Remove with qwen extensions uninstall ai-agent-skills. More in docs/qwen-code-setup.md.

Grok Build```bash
grok plugin install ysskrishna/ai-agent-skills --trust


Check with `grok plugin list`. Update with `grok plugin update ai-agent-skills`. Remove with `grok plugin uninstall ai-agent-skills`. Install directly: the marketplace route does not work for this repo. More in docs/grok-setup.md.

OpenCode

Add the plugin to `opencode.json` and restart OpenCode (npm or git):```json
{
  "plugin": ["@ysskrishna/ai-agent-skills"]
}
``````json
{
  "plugin": ["ai-agent-skills@git+https://github.com/ysskrishna/ai-agent-skills.git"]
}

OpenCode 2.0.4 or later uses the key plugins instead of plugin. Or install globally: opencode plugin @ysskrishna/ai-agent-skills --global. Check by asking the agent to list skills. OpenCode also reads .agents/skills, so npx skills add ysskrishna/ai-agent-skills works without a plugin. More in docs/opencode-setup.md.

Pi```bash
pi install npm:@ysskrishna/ai-agent-skills
``````bash
pi install git:github.com/ysskrishna/ai-agent-skills


Check with `pi list`. Update with `pi update`. Remove with `pi remove npm:@ysskrishna/ai-agent-skills` or `pi remove git:github.com/ysskrishna/ai-agent-skills`. More in docs/pi-setup.md.

Kimi Code

Inside Kimi Code:```text
/plugins install https://github.com/ysskrishna/ai-agent-skills

Choose Trust and install, then run /reload. Manage it later in /plugins. More in docs/kimi-setup.md.

Hermes Agent

Use a skills tap. Hermes then treats each skill as a normal skill:```bash
hermes skills tap add ysskrishna/ai-agent-skills
hermes skills install ysskrishna/ai-agent-skills/five-whys


Repeat the install for each skill you want. Update with `hermes skills update`. Remove with `hermes skills uninstall five-whys`.

A plugin install also works (`hermes plugins install ysskrishna/ai-agent-skills --enable`), but Hermes does not advertise plugin skills to the model, so use the tap. More in docs/hermes-setup.md.

Devin CLI```bash
devin plugins install ysskrishna/ai-agent-skills

Check with devin plugins info ai-agent-skills. Plugin commands need a Devin login. More in docs/devin-setup.md.

Muse```bash
git clone https://github.com/ysskrishna/ai-agent-skills.git
MUSE_EXPERIMENTAL_PLUGINS=1 muse plugins install ./ai-agent-skills --scope user
MUSE_EXPERIMENTAL_PLUGINS=1 muse plugins approve ai-agent-skills


Check with `muse skills list --source plugin`. Update with `muse plugins update ai-agent-skills`. Remove with `muse plugins remove ai-agent-skills`. More in docs/muse-setup.md.

OpenClaw (ClawHub)

Install **all skills in one bundle** from the ClawHub package (OpenClaw bundle-plugin):```bash
clawhub package download @ysskrishna/ai-agent-skills --output ./thinking-methods

Browse with clawhub package inspect @ysskrishna/ai-agent-skills. Per-skill installs stay on ClawHub too, for example clawhub install critical-thinking (see the Skills table). Maintainer sync: python3 scripts/publish_clawhub.py package plan and package publish (staged bundle folder upload; add --dry-run to preview) after bumping the repo version; per-skill sync uses plan / publish.

ClawHub category is context in openclaw.plugin.json. Topics are set on publish (scripts/publish_clawhub.py, max five): ai-agent-skills, thinking-methods, agent-skills, cursor, claude-code. Per-skill publish sends browse categories agents and knowledge plus the bundle topic set (bundle openclaw.plugin.json category stays context). Use clawhub CLI ≥ 0.26.0 for --topics / --categories on skill and package publish.

skills.sh (npx skills)

skills.sh installs skills into each agent's own folder. It supports Claude Code, Codex, Cursor, Gemini CLI, Windsurf, Antigravity, OpenClaw, GitHub Copilot, and many more.```bash

Install all skills from this repo

npx skills add ysskrishna/ai-agent-skills

List available skills

npx skills add ysskrishna/ai-agent-skills --list

Or install individual skills (--skill names match plugin / directory names)

npx skills add ysskrishna/ai-agent-skills --skill thinking-method-selector
npx skills add ysskrishna/ai-agent-skills --skill five-whys
npx skills add ysskrishna/ai-agent-skills --skill analytical-thinking
npx skills add ysskrishna/ai-agent-skills --skill systems-thinking
npx skills add ysskrishna/ai-agent-skills --skill first-principles-thinking
npx skills add ysskrishna/ai-agent-skills --skill design-thinking
npx skills add ysskrishna/ai-agent-skills --skill swot-analysis
npx skills add ysskrishna/ai-agent-skills --skill tradeoff-analysis
npx skills add ysskrishna/ai-agent-skills --skill prioritization
npx skills add ysskrishna/ai-agent-skills --skill strategic-thinking
npx skills add ysskrishna/ai-agent-skills --skill fermi-estimation
npx skills add ysskrishna/ai-agent-skills --skill critical-thinking
npx skills add ysskrishna/ai-agent-skills --skill pre-mortem
npx skills add ysskrishna/ai-agent-skills --skill ethical-thinking
npx skills add ysskrishna/ai-agent-skills --skill creative-thinking
npx skills add ysskrishna/ai-agent-skills --skill lateral-thinking
npx skills add ysskrishna/ai-agent-skills --skill six-thinking-hats


GitHub CLI (gh skill)

Requires GitHub CLI v2.90.0 or later.```bash
# Browse this repo's skills interactively
gh skill install ysskrishna/ai-agent-skills

# Install specific skills directly
gh skill install ysskrishna/ai-agent-skills thinking-method-selector
gh skill install ysskrishna/ai-agent-skills five-whys
gh skill install ysskrishna/ai-agent-skills analytical-thinking
gh skill install ysskrishna/ai-agent-skills systems-thinking
gh skill install ysskrishna/ai-agent-skills first-principles-thinking
gh skill install ysskrishna/ai-agent-skills design-thinking
gh skill install ysskrishna/ai-agent-skills swot-analysis
gh skill install ysskrishna/ai-agent-skills tradeoff-analysis
gh skill install ysskrishna/ai-agent-skills prioritization
gh skill install ysskrishna/ai-agent-skills strategic-thinking
gh skill install ysskrishna/ai-agent-skills fermi-estimation
gh skill install ysskrishna/ai-agent-skills critical-thinking
gh skill install ysskrishna/ai-agent-skills pre-mortem
gh skill install ysskrishna/ai-agent-skills ethical-thinking
gh skill install ysskrishna/ai-agent-skills creative-thinking
gh skill install ysskrishna/ai-agent-skills lateral-thinking
gh skill install ysskrishna/ai-agent-skills six-thinking-hats

# Target a specific host and scope when needed
gh skill install ysskrishna/ai-agent-skills tradeoff-analysis --agent codex --scope user

gh skill installs to the correct skill directory for the selected host, including GitHub Copilot, Claude Code, Codex, Cursor, Gemini CLI and Factory Droid.

Available skills

Start here

Skill Use it when Registry
thinking-method-selector You have a problem or decision but no method in mind. Picks one and runs it.

Understand a problem

Skill Use it when Registry
five-whys Something failed or keeps recurring and you need the real cause.
analytical-thinking A number moved or a question needs a driver tree, hypotheses and evidence.
systems-thinking A change may have side effects, or a problem keeps coming back across teams or services.
first-principles-thinking You want to test an inherited default or a copied approach from scratch.
design-thinking You need to understand users before designing or building.

Assess and decide

Skill Use it when Registry
swot-analysis You want to assess where a team, product or company stands, then act on it.
tradeoff-analysis You must pick between concrete options (A or B, build vs buy).
prioritization There is more work than capacity and you need an order and a cut line.
strategic-thinking You must choose a direction under constraints.
fermi-estimation You need a rough number fast (capacity, cost, size).

Check and de-risk

Skill Use it when Registry
critical-thinking You want a claim, plan or decision checked for weak spots.
pre-mortem Before a launch or commitment: imagine it failed and find out why.
ethical-thinking A plan could harm or unfairly burden people.

Generate ideas

Skill Use it when Registry
creative-thinking You want many fresh options before choosing.
lateral-thinking Ideas feel stuck or incremental and you need a different angle.

See it from several angles

Skill Use it when Registry
six-thinking-hats You want facts, feelings, risks, benefits and new ideas examined separately.

The full trigger text for each skill is in its SKILL.md frontmatter.

Usage

Describe your situation and the agent picks a matching skill from the descriptions, or name a skill directly.```text
We have to pick a job queue: Postgres, Redis or a managed service. Team of four.

What could go wrong with the database cutover next Friday?

Why did last night's deploy fail? Keep asking why until we reach something we can fix.

I have a pile of problems and no idea how to approach them.
```

To call a skill by name, use /skill-name after a skills.sh or gh skill install (for example /pre-mortem). Claude Code plugin installs namespace skills by plugin name (/:).

Built and maintained by

Thinking Methods is built and maintained by:

Name GitHub Role
Y. Siva Sai Krishna @ysskrishna Creator

Changelog

See CHANGELOG for release history.

Support

If you find this library helpful:

  • ⭐ Star the repository
  • 🐛 Report issues
  • 🔀 Submit pull requests
  • 💝 Sponsor on GitHub

License

MIT © Y. Siva Sai Krishna — see LICENSE for details.


Author's GitHub •
Author's LinkedIn •
Author's site •
ClawHub •
Report Issues

常见问题

8 问
答案由本站依据条目字段整理,不引用上游原文
Thinking Methods 怎么安装?

在客户端按下面的安装命令接入:npx skills add ysskrishna/ai-agent-skills。装完在客户端配置里引用同一个包或端点。

Thinking Methods 怎么在 Claude、Cursor、VS Code 里配置?

它的传输方式是stdio · 本地进程,在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,配置才会生效。

Thinking Methods 收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于服务本身和背后的接口额度,正式使用前按官方说明确认。

Thinking Methods 需要 API Key 吗?

资料标注需要会话登录认证,使用前准备好对应凭据,并按最小权限配置。

Thinking Methods 支持哪些客户端?

资料标注支持 Claude、Cursor、VS Code、Windsurf、Gemini CLI、Codex;其他客户端只要能配置 MCP Server,也能用同样的命令或地址接入。

Thinking Methods 能提供哪些能力?

资料显示能力标签包括 生产力。具体工具清单以官方文档为准。

Thinking Methods 需要什么运行环境?

属于本地进程,需要在本机装好运行环境。

Thinking Methods 是本地还是远程?

它是本地进程(stdio),由客户端在本机拉起。传输方式为stdio · 本地进程。

热门专题

按标签挑服务器
全部专题 →

服务器信息以作者仓库与官方文档为准;分类、传输方式与鉴权方式的中文说明为中文整理,最近核验 2026-11-20。