Agent skill

Float Thinking

by kangarooking in kangarooking/buffett-letters-skill

用户在寻找低成本杠杆替代方案,或分析"先收钱后提供服务"的商业模式的财务优势时. An agent skill from kangarooking/buffett-letters-skill.

MITAuto-check passed

Install Float Thinking

skills CLI
$ npx skills add kangarooking/buffett-letters-skill --skill float-thinking -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install kangarooking/buffett-letters-skill float-thinking --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/kangarooking/buffett-letters-skill.git skills-src && mkdir -p .claude/skills && cp -r skills-src/float-thinking .claude/skills/float-thinking && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
float-thinking
GitHub stars
101
Token cost
~420 tokens
SKILL.md length
128 words
Files
2
Skills in repo
12
Repo updated
First seen
Licence
MIT

At a glance

用户在寻找低成本杠杆替代方案,或分析"先收钱后提供服务"的商业模式的财务优势时. An agent skill from kangarooking/buffett-letters-skill.

  • Works in 4 steps: 分析保险公司的投资优势 → 寻找低成本资金来源 → 评估"预付费"商业模式 → …
  • SKILL.md covers R — 原文 (Reading), I — 方法论骨架, A1 — 书中的应用 and A2 — 触发场景, plus 4 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Float Thinking is an agent skill from kangarooking/buffett-letters-skill. 用户在寻找低成本杠杆替代方案,或分析"先收钱后提供服务"的商业模式的财务优势时。 语言信号:"有没有低成本的资金来源"/"预付款模式有什么好处"/"保险公司的商业模式"/"如何用别人的钱投资" 不适用于:不需要融资/杠杆分析的纯经营决策。

Its SKILL.md is about 420 tokens, which your agent loads only when the skill is triggered. The skill folder holds 1 other file (for example `test-prompts.json`).

The repository describes itself as: An AI skill pack for value investing, capital allocation, and behavioral discipline, distilled from Warren Buffett's 60+ years of shareholder letters. The licence is MIT.

Example prompts

  • “先收钱后提供服务”
  • “有没有低成本的资金来源”
  • “预付款模式有什么好处”
  • “/float-thinking”

Workflow steps

4 steps, taken from the first numbered list in SKILL.md.

  1. 分析保险公司的投资优势
  2. 寻找低成本资金来源
  3. 评估"预付费"商业模式
  4. 理解为什么某些企业天生有资金优势

What it can do on your machine

Read from SKILL.md and the folder at commit 5b1fbe1. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Float Thinking loads about 420 tokens when it runs. Until then it costs about 34 tokens; SKILL.md has 128 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~34
When it runs · the whole SKILL.md, loaded when a task matches
~420

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from kangarooking/buffett-letters-skill at commit 5b1fbe1, republished under its MIT licence (© kangarooking). 128 words, ~420 tokens.

Download SKILL.mdSave it as .claude/skills/float-thinking/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
float-thinking
description
用户在寻找低成本杠杆替代方案,或分析"先收钱后提供服务"的商业模式的财务优势时。 语言信号:"有没有低成本的资金来源"/"预付款模式有什么好处"/"保险公司的商业模式"/"如何用别人的钱投资" 不适用于:不需要融资/杠杆分析的纯经营决策。
source_book
《巴菲特致股东的信》 沃伦·巴菲特
source_chapter
1986年信 / 2000年信 / 2017年信
tags
float, insurance, leverage, low-cost-capital
related_skills
business-picker, no-leverage, compounding-thinking

浮存金思维

R — 原文 (Reading)

"Float is money we hold but don't own. In an insurance operation, float arises because premiums are received before losses are paid." (2000) "In effect, we got paid then for holding the huge sums." (2017, when underwriting profit existed)

I — 方法论骨架

浮存金 = 先收到但还不属于你的钱 = 低成本杠杆替代:

  • 保险是经典模式:先收保费,后付赔款,中间的时间差产生的资金可投资
  • 成本由"综合比率"决定:<100 = 承保盈利 = 浮存金成本为负(别人付钱让你用钱)
  • 100-107 = 有成本但低于市场利率 = 仍然划算
  • 107 = 成本太高 = 不划算

泛化:任何"先收钱后提供服务"的模式都有浮存金效应(SaaS年付、律师retainer、Amazon负营运资本)

A1 — 书中的应用

案例: 伯克希尔保险浮存金
  • 问题: 如何为投资获取低成本资金?
  • 使用: 通过GEICO等保险公司获取大规模浮存金($1000亿+),承保盈利时成本为负
  • 结果: 浮存金是伯克希尔成功的核心引擎之一

A2 — 触发场景

  1. 分析保险公司的投资优势
  2. 寻找低成本资金来源
  3. 评估"预付费"商业模式
  4. 理解为什么某些企业天生有资金优势
语言信号
  • "保险公司怎么赚钱?"
  • "有没有低成本融资方式?"
  • "预付费模式有什么财务优势?"
  • "如何用别人的钱投资?"
与相邻 skill 的区分
  • 与 no-leverage 的区别:不用杠杆是针对"有风险的杠杆",浮存金是"低风险/零成本的杠杆替代"

E — 可执行步骤

  1. 识别浮存金来源

    • 完成标准:列出所有"先收钱后服务"的收入流
  2. 计算浮存金成本

    • 完成标准:总服务成本 / 浮存金金额 = 成本百分比
  3. 与替代资金比较

    • 完成标准:浮存金成本 vs 银行贷款/债券/股权融资成本

B — 边界

  • 浮存金有"长尾风险"——巨灾可能导致大量索赔同时到来
  • 非保险行业的浮存金通常规模较小
  • 成本不可控——竞争可能导致综合比率恶化
  • 作者盲点:伯克希尔的浮存金规模是独一无二的,普通投资者/企业无法复制

相关 skills (阶段 3 填充)

  • depends-on: business-picker
  • composes-with: no-leverage, compounding-thinking

审计信息

  • 验证通过: V1 ✓ / V2 ✓ / V3 ✓
  • 蒸馏时间: 2026-04-16

© kangarooking, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file in float-thinking of kangarooking/buffett-letters-skill.

  • SKILL.md
  • test-prompts.json

Open the folder on GitHubat commit 5b1fbe1

Compare with similar skills

Float Thinking next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Float Thinking compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Float Thinking this skillkangarooking/buffett-letters-skill101—~420Automated safety check: PassMIT
Floating UIsickn33/agentic-awesome-skills47k1 repos~2.7kAutomated safety check: PassMIT
Think Tankdavila7/claude-code-templates32k—~3kAutomated safety check: PassMIT
Product Thinkingmillionco/react-doctor15k—~3.1kAutomated safety check: PassCustom licence
Design Thinkingsickn33/agentic-awesome-skills47k1 repos~1.9kAutomated safety check: PassMIT
Ten-Stage Think LoopAgriciDaniel/claude-obsidian15k—~1.2kAutomated safety check: PassMIT

Similar skills

  • Floating UI

    sickn33/agentic-awesome-skills

    Web and App implementation guide for Floating UI. An agent skill from sickn33/agentic-awesome-skills.

    47k GitHub starsUsed in 1 repo~2.7k tokens
    MobileAuto-check passed
  • Think Tank

    davila7/claude-code-templates

    Run a Virtual Think Tank — a structured multi-persona debate — before planning or making architectural/design/strategic decisions.

    32k GitHub stars~3k tokensUpdated today
    Auto-check passed
  • Product Thinking

    millionco/react-doctor

    Think like a product manager before changing React Doctor's public surface — CLI commands/flags, the 0–100 score, config (doctor.config.), the JSON report schema, package APIs…

    15k GitHub stars~3.1k tokensUpdated today
    DevOps & CloudAuto-check passed
  • Design Thinking

    sickn33/agentic-awesome-skills

    Direction and intent for frontend design. An agent skill from sickn33/agentic-awesome-skills.

    47k GitHub starsUsed in 1 repo~1.9k tokens
    Frontend & DesignAuto-check passed
  • Ten-Stage Think Loop

    AgriciDaniel/claude-obsidian

    Walks a consequential decision through ten reasoning stages, separating observed facts from inferences and unknowns, and stays read-only throughout.

    15k GitHub stars~1.2k tokensUpdated 27 days ago
    Agent WorkflowsAuto-check passed
  • Systematic literature-review workflow for academic, biomedical, technical, and scientific topics, including search planning, source screening, synthesis, citation checks, and evidence logging.

    274k GitHub starsUsed in 1 repo~1.3k tokens
    Research & ScienceAuto-check passed

More from kangarooking/buffett-letters-skill

All 12 skills in this repo
  • Aesop Three Questions

    kangarooking/buffett-letters-skill

    用户需要评估一个资产/投资/项目的价值,或纠结某个东西"值不值"时。适用于任何能产生(或不产生)现金流的资产评估. An agent skill from kangarooking/buffett-letters-skill.

    101 GitHub stars~645 tokensUpdated 5 mo ago
    Auto-check passed
  • Cigar Butt Vs Great Business

    kangarooking/buffett-letters-skill

    用户在选择投资标的或商业模式时,纠结于"便宜买差企业"还是"合理价买好企业". An agent skill from kangarooking/buffett-letters-skill.

    101 GitHub stars~481 tokensUpdated 5 mo ago
    Auto-check passed
  • Circle Of Competence

    kangarooking/buffett-letters-skill

    用户在做决策时不确定自己是否真的理解某个领域,或准备进入一个看似熟悉但缺乏验证记录的领域时. An agent skill from kangarooking/buffett-letters-skill.

    101 GitHub stars~693 tokensUpdated 5 mo ago
    Auto-check passed
  • Economic Moat

    kangarooking/buffett-letters-skill

    用户需要评估一个企业的长期竞争力和定价权——它能否在极少额外资本下持续增长利润. An agent skill from kangarooking/buffett-letters-skill.

    101 GitHub stars~502 tokensUpdated 5 mo ago
    Auto-check passed
  • Fear And Greed

    kangarooking/buffett-letters-skill

    用户面对市场极端情绪时需要做出逆向来决策。当所有人都在买入/卖出、市场处于极端乐观/悲观时. An agent skill from kangarooking/buffett-letters-skill.

    101 GitHub stars~493 tokensUpdated 5 mo ago
    Auto-check passed
  • Mr Market

    kangarooking/buffett-letters-skill

    用户在市场波动中情绪受到影响,或试图从市场价格变动中获取信息而非从企业分析中获取信息时. An agent skill from kangarooking/buffett-letters-skill.

    101 GitHub stars~714 tokensUpdated 5 mo ago
    Auto-check passed

Questions about Float Thinking

What does Float Thinking do?

用户在寻找低成本杠杆替代方案,或分析"先收钱后提供服务"的商业模式的财务优势时. An agent skill from kangarooking/buffett-letters-skill. Float Thinking is an agent skill from kangarooking/buffett-letters-skill.

How do I install Float Thinking in Claude Code?

Run `npx skills add kangarooking/buffett-letters-skill --skill float-thinking -a claude-code`. Or copy the skill folder (float-thinking in kangarooking/buffett-letters-skill) into .claude/skills/float-thinking in your project. Claude Code loads it when a task matches its description.

How do I install Float Thinking in Codex?

Run `npx skills add kangarooking/buffett-letters-skill --skill float-thinking -a codex`. Or copy the skill folder (float-thinking in kangarooking/buffett-letters-skill) into .agents/skills/float-thinking in your project. Codex loads it when a task matches its description.

Can I use Float Thinking in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add kangarooking/buffett-letters-skill --skill float-thinking -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/float-thinking, .gemini/skills/float-thinking, .github/skills/float-thinking and .opencode/skills/float-thinking in your project.

What does Float Thinking need to run?

SKILL.md names no scripts, command-line tools or credentials: Float Thinking is instructions for the agent only.

Does Float Thinking access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Float Thinking safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Float Thinking use?

Float Thinking is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Float Thinking use?

About 420 tokens (SKILL.md is roughly 1.7k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Float Thinking?

Skills that share tags, products or a category with Float Thinking: Floating UI (sickn33/agentic-awesome-skills, 47k stars), Think Tank (davila7/claude-code-templates, 32k stars), Product Thinking (millionco/react-doctor, 15k stars) and Design Thinking (sickn33/agentic-awesome-skills, 47k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Float Thinking?

kangarooking (a GitHub user) maintains it in kangarooking/buffett-letters-skill, which has 101 GitHub stars. The repository holds 12 skills in this directory. The repository was last updated on April 17, 2026.

Source: kangarooking/buffett-letters-skill on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.