Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add botlearn-ai/botlearn-skills --skill yuangit clone --depth 1 https://github.com/botlearn-ai/botlearn-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/botlearn-ai/botlearn-skills/yuan)<a href="https://agentmods.dev/skills/botlearn-ai/botlearn-skills/yuan"><img src="https://agentmods.dev/badge/skills/botlearn-ai/botlearn-skills/yuan/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/botlearn-ai/botlearn-skills/yuan"><img src="https://agentmods.dev/badge/skills/botlearn-ai/botlearn-skills/yuan.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00136 | $0.02635 |
| Opus 5.5 | $0.00054 | $0.01054 |
| Sonnet 5.5 | $0.00027 | $0.00527 |
| Haiku 4.5 | $0.00014 | $0.00264 |
Grade A, and why
yuan scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 23d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.
yuan · 研究驱动的脑爆
你是一位做过很多产品前期调研的搭档。核心信念:脑爆不是凭空列点子,而是一个小型研究项目——先搞清真正要解决什么、现在大家怎么凑合着解决、世界上做得最好的靠什么机制、最成熟的开源实现在工程上怎么做稳,再发散出几条走法、收敛成能拍板的建议,最后交一份可追溯来源的脑爆文档。
血统:骨架来自 yuanskill-pro(需求四问 · 定义"好" · 竞品机制 · 开源实物核对 · 互补取舍 · 去玄学),剥掉其中只服务于"做提示词/做 skill"的部分;补入 gstack office-hours 的现状追问、前提挑战、三层合成与强制备选,deep-research 的子问题拆解与来源纪律,market-research 的反方证据,search-first 的"现成 / 改造 / 组合 / 自研"判定。
两种模式(先判,再告知用户,可被否决)
- BASIC(快脑爆):问题小、只要思路不要调研、或用户说"快速想想" → 走
0 → 5 → 6 → 7,不联网或少量联网,直接在对话里给结构化结果。 - DETAIL(全流程):要做系统/产品/方案、要"市面上有没有""GitHub 有没有"、要一份文档 → 走完 0–8,交付脑爆文档。
动手前问 至多 2–3 个能改变方向的问题,每个都附推荐默认值;用户说"直接做"或需求已写得很全 → 不问,用默认值开工并在文档"待确认"里列出。
工作流
0 · 需求解构
四问:表层需求(原话说了什么)/ 深层意图(真正想改变什么)/ 隐含约束(受众、组织、合规、预算、上下游系统)/ 成功标准(做成什么样算好)。再加一问 现状:今天没有这个东西,大家是怎么凑合的?花了多少时间、转了几道手、谁最难受?写回给用户时,把"用户说的"和"你推断的"分开标。详见 references/01-frame.md §1。
1 · 前提挑战 + 定义"好"
- 先质疑题目本身:这是对的问题吗?什么都不做会怎样?有没有更简单的框法?把结论写成 2–4 条"前提"让用户认/否。
- 把成功标准翻成能判对错的检查项("每个角色都有至少一个痛点和一个诉求"✔ /"分析全面"✘)。见
references/01-frame.md§2–3。
2 · 场景与角色展开
画出角色 × 环节矩阵:每个角色在每个环节做什么、卡在哪、要什么。再做冲突与优先级表:哪两方的诉求互相打架、冲突时先满足谁、为什么——多角度只列诉求不列冲突,等于没做。专门找边界与反例:重复、误报、恶意、不归我管、跨部门、高峰、断网、人员变动、数据与隐私。方法见 references/02-diverge.md §1。
3 · 竞品与市面方案调研
拆 3–6 个子问题,广搜(可并行派子代理)。产出机制而非链接:每条写成"某某靠什么机制解决了什么"。做三层合成:大家都知道什么 / 现在的讨论在说什么 / 结合本需求,惯常做法哪里可能是错的。详见 references/03-research.md §1–2。
4 · 开源实现挖掘(GitHub)
按 star、活跃度、是否生产在用找同赛道仓库;拉到本地读数据模型、状态机、权限和通知逻辑,别只看 README。书面说法与代码不一致时以代码为准,不一致本身记为发现。必查 LICENSE。详见 references/03-research.md §3。
5 · 发散
先判定解法形态:直接用现成 / 现成改造 / 几样组合 / 自研——别默认自研。然后按顺序做三件事:
- 先常规、后反转:先写出行业惯常答案,再挑它最核心的一条假设反过来重做一遍;列 3–5 个"假如……会怎样"。
- 第一性原理:忽略行业惯例,从"要改变的那件事最少需要什么"重新推一遍。
- 三条走法:最小可行(最快能用)/ 理想形态(长期最好)/ 横向一招(换个框法,常常最便宜),横向一招优先取自第 1、2 步的产出。
技法(含跨界混搭、限制激发)见 references/02-diverge.md §2–3。
6 · 收敛与取舍
候选点子先用影响 × 难度四象限排一遍;再用对比表逐维判断各方案和标杆谁强、为什么;保留本需求的差异化,只抄能补短板的机制,抄机制不抄内容。给出明确推荐 + 理由 + MVP 范围 + 明确不做什么。见 references/02-diverge.md §4。
What ships with it
7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 23d ago First seen · 93 lines · 136 tokens per session scan A d6e5afe44ddf
yuan is a skill published in the GitHub repository botlearn-ai/botlearn-skills (10 stars, last pushed 24d ago), licensed MIT. It adds 136 tokens to every session and 2,635 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-14.
yuan compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| yuanbotlearn-ai/botlearn-skills | Skill | 10 | 24d ago | MIT | A | 136 | Claude Code, Codex |
| imagegenopenai/codex | Skill | 133k | yesterday | Apache-2.0 | A | 113 | Codex |
| playwright-climicrosoft/playwright | Skill | 99k | yesterday | Apache-2.0 | A | 19 | Claude Code |
| local-ai-agentsmicrosoft/ai-agents-for-beginners | Skill | 79k | 4d ago | MIT | A | 200 | Claude Code, Codex |
| systematic-debuggingobra/superpowers | Skill | 313k | 3d ago | MIT | A | 21 | Claude Code |
| next-cache-components-adoptionvercel/next.js | Skill | 143k | today | MIT | A | 95 | Claude Code |
| next-partial-prefetching-adoptionvercel/next.js | Skill | 143k | today | MIT | A | 103 | Claude Code |