RadarTopicsBuildersWeeklyReads
Open Source Radar
HughYau/

qiushi-skill

GitHubWebsite

qiushi-skill is a JavaScript-based collection of AI agent skills derived from dialectical materialism and practical philosophy, offering nine methodological tools and multiple installation methods for integrating into AI workflows.

3.7kstars
277forks
20issues
MITlicense
2026since
Star historydaily snapshots by VibeCrowd

Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).

Alternatives & relatedmatched by topic overlap
Reviewgenerated from repository data · Aug 5, 2026

What it is

qiushi-skill is an AI Agent Skills collection distilled from classical dialectical materialism and practical philosophy. It provides a total principle and nine tooling methods to arm AI reasoning processes. Each skill references original texts in its directories.

How it works

The project structures nine skills under skills/, plus a workflows component for coordinating method calls. It includes prompts and reference guides to support actionable implementation. The README describes a layering: a general principle (实事求是), then philosophical base (矛盾分析法, 实践认识论), followed by work methods (调查研究, 群众路线, 批评与自我批评), and strategic/operational guidance (持久战略, 集中兵力, 星火燎原, 统筹兼顾).

Getting started

There are several installation options:

  • "npx qiushi-skill" for interactive installation. The README shows:
npx qiushi-skill
  • Non-interactive installations via npx qiushi-skill install with various targets and scopes, e.g.:
npx qiushi-skill install --target claude-code --scope user
  • Claude Code official Marketplace installation via marketplace.json and commands:
/plugin marketplace add HughYau/qiushi-skill
/plugin install qiushi-skill@qiushi-skill
  • Claude Plugin Hub alternative:
npx claudepluginhub hughyau/qiushi-skill
  • Direct AI agent installation instructions are provided as a multiline block beginning with:
请帮我安装 qiushi-skill:

The README also provides verification steps:

npx qiushi-skill validate
bash tests/validate.sh
powershell -NoLogo -NoProfile -ExecutionPolicy Bypass -File tests/validate.ps1

Recent releases

The Releases section lists latest as 0 with:

  • none

Traction

The repository shows: stars: 3674, forks: 277, open_issues: 20. The README includes a Star History section but no separate traction text.

Behind the repo

The repository links to related project AcademicForge in the description and references Claude Code marketplace integration via hidden paths under .claude-plugin/. There is no explicit company or startup link beyond the author and related project suggestion.

Caveats

Licensing is MIT. The repository is created on 2026-03-25 and last_push on 2026-07-20. The README describes many platforms and install paths (.codex, .opencode, .openclaw, .hermes, .nanobot) with corresponding INSTALL.md files.

Agent skills inside · 11
All skills →
contradiction-analysis触发:当问题复杂、存在多个冲突因素、优先级不清,或你不知道应该先解决什么时调用;常见信号包括 trade-off、瓶颈、根因不明、主次不清、多个问题互相牵制。Otherinvestigation-first触发:当你准备下判断、做决策或提出建议,但事实、上下文或一手信息还不充分时优先调用;常见信号包括 unknowns、信息缺口、证据不足、领域陌生、需要先摸清现状。Otherprotracted-strategy触发:当目标长期、任务复杂、资源暂时处于劣势,或短期无法速胜但又不能放弃时调用;常见信号包括 long-term effort、phased plan、endurance、战略耐心、需要分阶段推进。Workflow & Productivitymass-line触发:当你需要收集多方意见、把零散反馈整合成可执行方案,或把方案带回真实使用者/执行者验证时调用;常见信号包括 stakeholder input、user feedback、意见汇总、对齐与验证。Othercriticism-self-criticism触发:当一项工作已经完成、进入阶段验收、收到批评反馈,或反复出现同类错误需要系统纠偏时调用;常见信号包括 review、audit、retrospective、quality check、纠错与复盘。Securityconcentrate-forces触发:当多个任务同时争夺时间、注意力、算力或预算,必须确定主攻方向并停止分散用力时调用;常见信号包括优先级过多、资源紧张、推进分散、需要决定先做什么。Other
SharePost on XLinkedIn
All trending reposRevenue-verified startups →