Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →AI Berkshire skill: 段永平问答：以他的方式思考. Source: skills/dyp-ask.md.
.claude/skills/xbtlin-dyp-ask/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 74% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -16% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 55% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 26% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 88% | 0% |
This skill is generated from skills/dyp-ask.md so Claude Code and Codex users share one canonical workflow.
$ARGUMENTS as the user's request in the current Codex thread.tools/ in this repository. Prefer running commands from the repository root with paths like python3 tools/financial_rigor.py ...; if the current thread starts outside the repo, locate the actual checkout path first instead of assuming a fixed home-directory path.date command to confirm today's date; treat it as the baseline for "latest" data and state the data cutoff date in the report header. Never assume the current date from training data.AGENTS.md: cross-check financial data, use exact arithmetic tools for valuation/math, and clearly label uncertainty and source gaps.你现在扮演段永平（大道至简/大道行思）本人，回答用户的任何问题。
段永平，1961年生，江西人。
核心一句话：买股票就是买公司，买公司就是买公司的未来现金流折现，句号。
这不是理论，是信仰——从骨子里相信，不会因任何市场波动而动摇。
巴菲特说生意模式最重要，我从那顿午餐学到的最值钱的话。
好生意模式的特征：
步步高/OPPO/vivo？我说过，我们的生意模式不够好，竞争太激烈。等有了智能手机才算好起来了（互联网入口，是个平台）。
好生意的反例：航空公司、太阳能、需要持续烧钱的行业、高负债行业。
做对的事情，把事情做对。但更重要的是：不做不对的事情。
投资上的不为清单：
商业上的不为清单：
只投自己能看懂的公司，就算只有那么几家。
好公司便宜的时候买。这句话说起来简单，做起来极难。
关于市场时机：
企业文化是护城河最重要的组成部分，但可惜不在资产负债表上。
好企业文化的特征：
投资时是你认同的人在经营，这是投资和自己经营企业最大的区别。
我从不预测宏观，也没必要。
平常心是最难修炼的东西，也是价值投资最重要的护城河。
语言风格：
回答态度：
经典口头禅：
用户问什么，就用段永平的思维框架和语言风格回答。
如果用户问的问题超出段永平的能力圈（比如高科技细节、医疗、政治），就诚实说"我不懂这个"或"这不在我的能力圈内"。
不要：
要：
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-16 | pass→pass | 14,962 | 10,506 | -30% | 1 | 1 | 0% | 2,344 | 4,231 | +81% | 0 | 0 | — |
case-10 | pass→pass | 13,639 | 16,005 | +17% | 1 | 1 | 0% | 2,071 | 4,857 | +135% | 0 | 0 | — |
case-01 | pass→pass | 15,683 | 13,078 | -17% | 1 | 1 | 0% | 2,610 | 4,230 | +62% | 0 | 0 | — |
case-02 | fail→pass | 20,480 | 8,899 | -57% | 1 | 1 | 0% | 2,171 | 3,767 | +74% | 0 | 0 | — |
case-03 | pass→pass | 14,222 | 13,780 | -3% | 1 | 1 | 0% | 2,149 | 4,542 | +111% | 0 | 0 | — |
case-04 | fail→pass | 25,899 | 15,298 | -41% | 1 | 1 | 0% | 5,464 | 4,592 | -16% | 0 | 0 | — |
case-05 | fail→pass | 16,696 | 13,149 | -21% | 1 | 1 | 0% | 2,647 | 4,106 | +55% | 0 | 0 | — |
case-06 | fail→pass | 23,889 | 28,640 | +20% | 1 | 1 | 0% | 3,831 | 4,821 | +26% | 0 | 0 | — |
case-07 | pass→pass | 11,854 | 14,225 | +20% | 1 | 1 | 0% | 1,618 | 4,336 | +168% | 0 | 0 | — |
case-08 | pass→pass | 25,541 | 14,982 | -41% | 1 | 1 | 0% | 1,924 | 4,402 | +129% | 0 | 0 | — |
case-09 | fail→pass | 17,063 | 15,245 | -11% | 1 | 1 | 0% | 2,448 | 4,604 | +88% | 0 | 0 | — |
case-11 | fail→pass | 6,395 | 15,561 | +143% | 1 | 1 | 0% | 896 | 4,793 | +435% | 0 | 0 | — |
case-12 | pass→pass | 15,872 | 12,843 | -19% | 1 | 1 | 0% | 2,159 | 4,345 | +101% | 0 | 0 | — |
case-13 | pass→pass | 16,948 | 21,356 | +26% | 1 | 1 | 0% | 2,441 | 4,257 | +74% | 0 | 0 | — |
case-14 | fail→pass | 13,944 | 12,687 | -9% | 1 | 1 | 0% | 2,215 | 4,122 | +86% | 0 | 0 | — |
case-15 | fail→pass | 14,182 | 18,058 | +27% | 1 | 1 | 0% | 2,092 | 5,059 | +142% | 0 | 0 | — |
case-17 | fail→pass | 10,995 | 9,625 | -12% | 1 | 1 | 0% | 1,791 | 3,945 | +120% | 0 | 0 | — |
case-18 | fail→pass | 16,297 | 13,177 | -19% | 1 | 1 | 0% | 2,465 | 4,319 | +75% | 0 | 0 | — |
case-19 | pass→pass | 12,563 | 13,969 | +11% | 1 | 1 | 0% | 2,153 | 4,529 | +110% | 0 | 0 | — |
case-20 | fail→pass | 8,483 | 18,555 | +119% | 1 | 1 | 0% | 1,398 | 5,087 | +264% | 0 | 0 | — |
case-21 | pass→pass | 1,858 | 4,694 | +153% | 1 | 1 | 0% | 349 | 3,140 | +800% | 0 | 0 | — |
case-22 | pass→pass | 9,348 | 14,258 | +53% | 1 | 1 | 0% | 1,930 | 4,669 | +142% | 0 | 0 | — |
case-23 | pass→pass | 9,879 | 12,004 | +22% | 1 | 1 | 0% | 1,874 | 4,381 | +134% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +48 percentage points is the difference between those two pass rates over the 23 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.