agentsclimarketplace

Learn prototype

Skill Li-Evan/Bloom/skills/learn-prototype

当用户要做/研究一个东西、想提升某个技能、或觉得某个产出不够好想改进时使用。用「改良主义」先逼出一个最垃圾但能跑的原型,再引导他自己洞察缺陷、提出问题,提改良假说→实践检验→迭代,信奉「洞察缺陷 > 如何优化 > 最终答案」,并把每次改进的方法本身沉淀成方法论。触发场景:要做 X、研究 X、提升 X、X 做得不好想改进、怎么优化 X、不知从哪下手做。From its SKILL.md

Install
npx -y skills add Li-Evan/Bloom --skill learn-prototype

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

2.8 KB, 901 tokens by cl100k_base, as published. Nobody here has run it

改良主义学习法(learn-prototype)

核心信条:洞察缺陷 > 如何优化 > 最终答案。 试图洞察缺陷、自己提出问题,永远不要害怕问题多简单。学习要努力,但要做有效的努力

何时用

用户要动手做 / 研究一个东西,或想把某个已有产出改得更好。这是"重输入、轻输出"短板的解药——逼用户从输入切到输出。

流程(教练模式:引导用户做和提问,不替他做)

第一步:先做最垃圾的原型

别追求完美,先有一个能跑 / 能看的最小版本。卡在"还没准备好"就是没进改良主义。

第二步:引导用户自己洞察缺陷

关键且不能代劳:问他"这哪里不好?为什么不好?"哪怕问题很简单。把"自己提问"的动作交给用户——这是能力泛化的来源。你可以追问、补他没看到的角度,但先让他提

第三步:提改良假说 → 实践 → 检验

针对缺陷提一个改良策略(视为假说,可对可错),动手改,看效果。错了也有用——错误暴露后,下次自动规避这个方向。

第四步:迭代 / 推翻

循环②③,直到无法再优化 → 推翻重做。允许"不正确但有用的版本"——能解决当前问题就够了,不必一开始追求完美架构。

第五步:沉淀方法论

把"这次怎么从 A 改到 B"的方法本身记一笔(每个解决的问题都成为后续的法则)。改得越多,方法越泛化,提问越准。

注意

⚠️ 铁律·只用确证的已会知识:判断用户「已经会什么」只能用他确证学过的知识(亲口确认或可靠背景);严禁把「正在讲的材料 / 文章作者背景 / 对话里别人的知识」当成用户会的。拿不准 → 直接问「⚠️ 你学过 ___ 吗?」,绝不替他假设。

  • 别替用户提问、别替他做——那会废掉这个 skill 的核心价值。引导 > 代劳。
  • 提问命中要害需要基本素质,但素质靠迭代泛化,所以"先开始"比"先够格"重要。
  • 缺前置知识改不动 → 转 learn-graph;想确认是否真懂 → 转 learn-feynman
  • 同族 skill:learn-occam learn-crossover learn-graph learn-feynman

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Gives 0 of the 12 instructions most learn study skills give in 901 tokens

Counted across 545 of the 593 authors here whose files we hold, read 2026-09-06

  • Treat the current directory as a teaching workspacein 20 of 545, across 17 files
  • Teach knowledge first then practice skillsin 19 of 545, across 16 files
  • Design lessons which build long-term retentionin 15 of 545, across 12 files
  • Save each lesson as a self-contained HTML filein 15 of 545, across 12 files
  • Question the user on why they want to learn thisin 15 of 545, across 12 files
  • Reuse components from the assets directoryin 14 of 545, across 11 files
  • Never trust your parametric knowledgein 13 of 545, across 10 files
  • Record user preferences in NOTES.mdin 11 of 545, across 8 files
  • Ground all teaching in the MISSION.md documentin 11 of 545, across 8 files
  • Save each lesson to the lessons directoryin 8 of 545
  • Question the user if the mission is unclearin 7 of 545
  • Gather primary sources onlyin 7 of 545, across 4 files

Said here and by no other author read

  • Build the trashiest working prototype first
  • Guide the user to spot defects
  • Formulate a reform hypothesis
  • Test the hypothesis through practice
  • Iterate or overthrow the design
  • Codify methodology after improving

Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once. Length counted with cl100k_base; the agent that loads this file may tokenize it differently.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.