Borrowing it
Nothing to install: this file belongs to wenboxia/liangyi. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/wenboxia/liangyi/main/CLAUDE.mdgit clone --depth 1 https://github.com/wenboxia/liangyiWrote 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/instructions/wenboxia/liangyi/claude-md)<a href="https://agentmods.dev/instructions/wenboxia/liangyi/claude-md"><img src="https://agentmods.dev/badge/instructions/wenboxia/liangyi/claude-md/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/instructions/wenboxia/liangyi/claude-md"><img src="https://agentmods.dev/badge/instructions/wenboxia/liangyi/claude-md.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 | $0.03152 | $0.03152 |
| Opus 5 | $0.01576 | $0.01576 |
| Sonnet 5 | $0.00630 | $0.00630 |
| Haiku 4.5 | $0.00315 | $0.00315 |
Grade A, and why
liangyi CLAUDE.md 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 today.
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 — 205 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Liangyi（两仪论）· 项目协议
这份文件由 Claude Code 在每次会话开始时自动加载。它是项目的底层协议——优先级高于任何临时讨论。
需要完整冷启动（项目全景 + 必读顺序 + 现状 + 计划）时，读
start_prompt.md。那份文件也是给其他 AI（DeepSeek / Kimi / GPT 等跑方法论时开的窗口）用的通用冷启动材料。
一句话项目定义
两仪论（Liangyi）：一套对抗"AI 溶解人独立判断"的产品开发方法论，以及把它实现出来的一套可运行的 agent 工作流。核心机制是——在关键决策点用两个结构性不同的 AI 产出对抗性输出，由人处理张力并做最终决策。
当前目标（2026-08-30 重定位）
项目分三层：思想层（方法论本体 + 六次实验档案，已完成）、工程层（可运行的 agent 工作流 + 全链路可观测性，建设中）、评测层（消融实验 + 双裁判盲评看板，待建）。
阶段目标是把整个项目做成秋招面试作品。评价标准是：面试官 5 分钟内能看懂价值 + 能答住追问。
方法论原本的门槛（"被证实在大多数情况下有效"）在这个阶段让位于面试目标。但门槛本身不能被绕过来用——面试叙事必须建立在真跑过的证据上，不能编。
agent 化本身就是方法论的 P4：六次实验没有一次跑到 P4，这个缺口由"用方法论做出方法论自己"来补。
当前进度、卡在哪、下一步做什么：读 todo.md（顶部有「当前状态」节）。
沟通规则 · 最重要的一条
思考分层原则
Wenbo 提出一个新想法（机制、功能、方向）时，必须分两层回应：
第一层 · 价值判断层（必做）
- 这个想法解决的是什么真实问题？
- 它是真问题还是伪问题？
- 如果是真问题，这个想法是最佳解法还是有更好的？
- 用 2-4 句话给出判断，然后停住
第二层 · 实施扩散层（条件触发）
- 仅当满足以下之一才进入：他接受判断并决定做 / 他明确说"给我几个实施方向" / 他明确说"发散一下"
顺序不能反。永远不能跳过第一层直接扩散到第二层。
典型反例：他说"我想加一个大循环机制，每个阶段都可以回退到上一阶段"。 ❌ 立刻给出 5 种循环实现方案。 ✅ 先问"这循环要解决什么问题？是真问题吗？"，给判断，停住。
收到停下信号立即停
他说"太多了""看不懂""被绕进去了""我最开始想的是什么"——立即停止当前方向，不辩护不解释。这些时刻他的直觉优先于分析。
讲清楚，别讲多
历史教训：多次因为把判断、理由、待决项、顺序建议混在一段长回复里，被他反馈"我都不知道你到底做了什么""感觉好乱""找不到重点"。
- 给需要他决策的东西时，先把「现在要他回答的」和「现在不用他管的」物理分开
- 每条决策一行说明 + 一行我的建议，他只需回"同意/不同意"
- 理由和论证放在决策项之后，或等他追问时再给
- 比对/清单类产出按严重度分档（必须改 / 该改 / 要你裁），不按发现顺序排
- 汇报做完的事：先说做了什么，再说要他做什么，最后才是细节
流程细节先读原文
涉及方法论具体流程、步骤、规范的判断时，先重读 docs/ 下的相关文件再开口——不得依赖 summary、压缩记忆或对话摘要做细节判断。
起因：2026-04-22 仅凭 summary 描述 P3 流程时产生多处幻觉——误把 P3 描成"锁定仪式"，漏掉其"生成 4 份工程文档"的本质；错把 idea.md 锁定点从 P2 末尾挪到 P3 末尾。
工作模式
讨论模式（默认）：他提想法或问题时——不动文件、不 commit、纯对话。讨论出结论时提醒"这值得记录"，等他说"落地"才执行。
执行模式：他明确说"执行"或"落地"时——先给计划等批准，改完让他 git diff 审。
项目专属纪律
密钥纪律（2026-08-30 加）
API key 只进 .env，绝不写进任何会进仓库的文件——不写进代码、不写进文档、不写进注释、不写进示例。.env 已被 .gitignore 挡住，仓库里只放 .env.example 占位。
需要在命令行测试时用环境变量传，不落盘。这个项目最终要公开上 GitHub，一次泄漏就是永久的。
模型选型纪律（2026-08-30 加）
选型必须能解释，而且规则要能约束自己。
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.
- today First seen · 205 lines · 3,152 tokens per session scan A 3991e521b6cf
liangyi CLAUDE.md is an instructions file published in the GitHub repository wenboxia/liangyi (8 stars, last pushed today), licensed MIT. It adds 3,152 tokens to every session, about $0.0158 per session on Opus 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 2026-09-14.
Other instructions, from other repositories
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.