hughyau

hughyau/qiushi-skill

11 resources in this repository

GitHub
๐ŸŽฏ11

๐ŸŽฏSkills11

๐ŸŽฏarming-thought๐ŸŽฏSkill

Entry skill of the **ๆฑ‚ๆ˜ฏ (Qiushi)** collection that auto-triggers at the start of each top-level conversation (but skipped by delegated sub-agents) to enforce Mao Zedong-derived "seek truth from facts" (`ๅฎžไบ‹ๆฑ‚ๆ˜ฏ`) rules โ€” fact-backed conclusions, verification before "done", explicit "I don't know" admissions, and lightweight routing to downstream qiushi skills (contradiction-analysis, practice-cognition, etc.) only when clearly useful.

arming-thought
๐ŸŽฏinvestigation-first๐ŸŽฏSkill

Qiushi (ๆฑ‚ๆ˜ฏ) methodology skill teaching **investigation first (่ฐƒๆŸฅ็ ”็ฉถ)** โ€” "without investigation there is no right to speak" โ€” requiring AI agents to gather facts before making claims, derived from *"Oppose Book Worship"* with the metaphor "investigation is like ten months of pregnancy, solving problems is like the birth."

investigation-first
๐ŸŽฏcontradiction-analysis๐ŸŽฏSkill

Qiushi (ๆฑ‚ๆ˜ฏ) methodology skill teaching **contradiction analysis (็Ÿ›็›พๅˆ†ๆžๆณ•)** โ€” identify contradictions, seize the principal contradiction, and distinguish the nature of contradictions to solve complex problems; derived from *"On Contradiction"* with the principle "grasp the main contradiction and all problems will be solved."

contradiction-analysis
๐ŸŽฏconcentrate-forces๐ŸŽฏSkill

Qiushi (ๆฑ‚ๆ˜ฏ) methodology skill teaching the **concentrate-forces (้›†ไธญๅ…ตๅŠ›)** principle โ€” "concentrate superior forces to destroy the enemy one by one" โ€” for AI agents to focus resources on one high-impact target at a time rather than spreading effort thin, derived from *"Problems of Strategy in China's Revolutionary War."*

concentrate-forces
๐ŸŽฏpractice-cognition๐ŸŽฏSkill

Qiushi (ๆฑ‚ๆ˜ฏ) methodology skill teaching **practice-cognition epistemology (ๅฎž่ทต่ฎค่ฏ†่ฎบ)** โ€” the "practice โ†’ cognition โ†’ practice again" spiral where truth is verified through action, not theory alone; derived from *"On Practice"* with the maxim "practice is the sole criterion for testing truth."

practice-cognition
๐ŸŽฏoverall-planning๐ŸŽฏSkill

Qiushi (ๆฑ‚ๆ˜ฏ) methodology skill teaching **overall planning (็ปŸ็ญนๅ…ผ้กพ)** โ€” mobilize all positive factors, reject one-sidedness, and seek dynamic balance when juggling competing objectives; derived from *"On the Ten Major Relationships"* for multi-goal decision-making and trade-off analysis.

overall-planning
๐ŸŽฏcriticism-self-criticism๐ŸŽฏSkill

Qiushi (ๆฑ‚ๆ˜ฏ) methodology skill teaching **criticism and self-criticism (ๆ‰น่ฏ„ไธŽ่‡ชๆˆ‘ๆ‰น่ฏ„)** โ€” "learn from past mistakes to prevent future ones, cure the sickness to save the patient" โ€” requiring systematic review of completed work, candid assessment of gaps, and iterative improvement; derived from *"On Coalition Government."*

criticism-self-criticism
๐ŸŽฏspark-prairie-fire๐ŸŽฏSkill

Qiushi (ๆฑ‚ๆ˜ฏ) methodology skill teaching the **spark-can-start-a-prairie-fire (ๆ˜Ÿ็ซ็‡ŽๅŽŸ)** principle โ€” establish base areas instead of roaming, start small, accumulate strength, and scale from a foothold rather than attempting everything at once; derived from *"A Single Spark Can Start a Prairie Fire."*

spark-prairie-fire
๐ŸŽฏworkflows๐ŸŽฏSkill

Qiushi (ๆฑ‚ๆ˜ฏ) **workflow orchestration layer** โ€” a cross-skill coordination skill that defines how to chain multiple methodology tools (contradiction analysis, investigation, mass line, etc.) together with calling order and data passing rules, serving as the routing backbone for the 9-tool methodology collection.

workflows
๐ŸŽฏprotracted-strategy๐ŸŽฏSkill

Qiushi (ๆฑ‚ๆ˜ฏ) methodology skill teaching **protracted strategy (ๆŒไน…ๆˆ˜็•ฅ)** โ€” "strategically despise the enemy, tactically take them seriously" โ€” for long-term complex tasks that require neither rushing nor giving up, but sustained, phased progress; derived from *"On Protracted War."*

protracted-strategy
๐ŸŽฏmass-line๐ŸŽฏSkill

Qiushi (ๆฑ‚ๆ˜ฏ) methodology skill teaching the **mass line (็พคไผ—่ทฏ็บฟ)** โ€” "from the masses, to the masses" โ€” a feedback loop of collecting scattered user/stakeholder input, systematizing it into actionable plans, returning it for validation, and iterating; derived from *"Some Questions Concerning Methods of Leadership."*

mass-line