# HughYau/qiushi-skill

Qiushi-Skill: Build agents that investigate first, focus on the main contradiction, validate in practice, and keep pushing until the work is actually done. 求是Skill——从经典唯物辩证法与实践哲学中提炼出一条总原则和九大方法论工具武装AI大脑。

Repository: https://github.com/HughYau/qiushi-skill
Canonical: https://ross.abutalabs.com/products/qiushi-skill
Homepage: http://hughyau.com/qiushi-skill/
Language: JavaScript
License: MIT
License Family: permissive
Topics: ai-agents, methodology, skills, workflow
Last push: 2026-07-20T19:04:56+00:00

## Health v2 (maintenance only)
Score: 56/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 93, release rhythm 35, longevity 11
- inputs: {"age_days": 161, "days_push": 44, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3735, forks 286 (observed 2026-08-28T04:08:15.792692+00:00)

## What it is
Qiushi-Skill is a collection of AI Agent Skills that distills a general principle and nine methodological tools from classical materialist dialectics and practice philosophy to structure how AI agents analyze and execute tasks. It is distributed as an npm package and installable into agent tools like Claude Code, Cursor, and Codex.

## Use cases
- make my AI agent investigate before answering
- help the agent identify the main contradiction in a complex problem
- add self-review and critique loops to agent workflows
- improve agent persistence on long-running tasks
- prioritize and focus agent effort on what matters
- structure long-term multi-stage task planning for agents

## When to choose
- you use Claude Code, Cursor, Codex, or similar agent tools and want a systematic reasoning methodology layered on top
- your agent gives premature answers without investigating context
- you want structured problem-analysis, prioritization, and iteration discipline for agent workflows

## When to avoid
- you expect the skill to generate answers or automate tasks directly - it only constrains the reasoning process
- you need a general agent framework or runtime rather than a methodology/prompt skill pack
- you prefer agents without opinionated philosophical workflow constraints

## Facets
- artifact type: plugin
- maturity: active
- function: agent-framework, prompt-engineering, workflow-automation, developer-tools
- domain: developer-tools
- platform: cross-platform, cli
- tags: ai-agent-skills, methodology, claude-code, cursor, dialectics, problem-solving, chinese, ai-agents, automation, nodejs

## Member repositories
- HughYau/qiushi-skill (main) score 56

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:15.792692+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:29:11.409470+00:00, confidence not recorded.
  - readme: https://github.com/HughYau/qiushi-skill (fetched 2026-08-28T04:08:15.792692+00:00, sha ff6965646df1)
  - homepage: http://hughyau.com/qiushi-skill/ (fetched 2026-08-29T09:23:59.466407+00:00, sha 2d47d82540cd)
  - registry_npm: https://registry.npmjs.org/qiushi-skill (fetched 2026-08-29T09:23:59.475749+00:00, sha ed524702a71b)
- Data as of 2026-08-30T08:39:29.467469+00:00.