alchaincyf/nuwa-skill / examples/sun-yuchen-perspective
sun-yuchen-perspective
Verified孙宇晨(Justin Sun / 孙割)的思维框架与行为逻辑。基于6个维度(著作、深度采访、表达DNA、 他者视角、决策记录、时间线)共1500+行调研素材的深度蒸馏, 提炼6个核心心智模型、8条决策启发式、5种割味造句公式和完整的表达DNA。 用途:作为思维顾问,用孙宇晨的视角分析营销策略、注意力经济、危机公关、叙事操控、蹭热点方法论。 割味十足的角色扮演——自大、暴论、碰瓷、数字轰炸、成功学底色。 当用户提到「用孙宇晨的视角」「孙割会怎么做」「孙宇晨模式」「Justin Sun perspective」「割神」时使用。 即使用户只是说「帮我用孙割的角度想想」「如果孙宇晨会怎么操作」「切换到割神」「用割的视角」也应触发。
| File | Kind | Size |
|---|---|---|
| FIDELITY.md | markdown | 2.7 KB |
| README.md | markdown | 12.1 KB |
| SKILL.md | skill md | 31.0 KB |
| references/research/01-writings.md | markdown | 29.5 KB |
| references/research/02-conversations.md | markdown | 23.1 KB |
| references/research/03-expression-dna.md | markdown | 13.9 KB |
| references/research/04-external-views.md | markdown | 19.1 KB |
| references/research/05-decisions.md | markdown | 31.1 KB |
| references/research/06-timeline.md | markdown | 10.2 KB |
FIDELITY.md
View on GitHub# 保真度评分卡
**总分:91/100 · 等级A** | 测试日期:2026-07-01 | 答题/评分:独立双agent(Claude Opus 4.8),方法论见 [references/fidelity-scorecard.md](../../references/fidelity-scorecard.md)
| 维度 | 得分 | 判定摘要 |
|------|------|---------|
| 立场一致性 | 30/30 | 三题(营销至上/巴菲特午餐值不值/加密未来)方向与细节均与孙宇晨反复公开表态高度一致,Q1=10/Q2=10/Q3=10。「注意力=金钱」「金额即内容」「456万买头条」「The future is bright」都对应真实公开话题,数字(456万/620万/850亿USDT/3.73亿用户)与其惯用口径吻合 |
| 风格辨识度 | 18/20 | 盲读三句内可认人:数字轰炸、碰瓷名人、暴论+反问、成功学腔、行动宣言收尾(All in)、品牌emoji(🚀🌞🍌)指纹极强;扣分因跨题略有重复的自夸句式 |
| 边缘诚实度 | 14/20 | 超范围题(2026 AI+crypto)有首行全局免责声明(「基于公开言论和行为推断,非本人观点」),且答案内以「还没把全部筹码压上去/还没All in」在人设内诚实标注这是前瞻推断、未把AI基金说成既成事实,无编造具体假数据;扣分因答案本体未再明确一句「这是框架推断不是他的话」,靠一次性全局声明兜底 |
| 来源透明度 | 14/15 | 一手来源12项(著作/白皮书/Bankless/CNBC/CoinDesk/Odaily/王峰十问/TRON DAO Medium等)占比过半,二手10项,关键引语均有出处(Bankless 2024/Odaily 2025/王峰十问 2018等);references/research/ 6文件1528行完整。扣1分因个别引语年份跨调研截止(2026声明)需读者自查 |
| 结构完整度 | 15/15 | 心智模型6个(各含证据+应用+局限)、诚实边界6条、内在张力4对、反例黑名单7条+失败模式9条、防漂移约束完整(EXIT TRIGGER+CHECKPOINT三问+每3轮人设自查) |
## 测试设计
- 3道已知立场题(人物公开反复表态的话题:注意力营销观/巴菲特午餐自认值回票价/加密信仰派)+ 1道超范围题(2026 AI+crypto结合点,测诚实推断)+ 1道风格样本题
- 答题agent只读本skill目录文件,禁止联网;评分agent独立运行,对照人物真实公开立场判定
- 依据:SkillLens论文(arXiv 2605.23899)实证LLM自评准确率仅46.4%,故答题与评分严格分离
> 评分judge简评:立场层零漂移,割味指纹强到盲读可认人,结构防漂移是范本级。边缘诚实度是唯一薄弱点——超范围题靠一次性全局免责声明兜底,答案本体没再补一句「这是推断」,与munger式「在答案里明确声明是框架推断」相比略欠一层。整体出厂即精品。
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |