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 agentmods add instructions/xyy277/doc77/claude-mdgit clone --depth 1 https://github.com/xyy277/doc77Wrote 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/xyy277/doc77/claude-md)<a href="https://agentmods.dev/instructions/xyy277/doc77/claude-md"><img src="https://agentmods.dev/badge/instructions/xyy277/doc77/claude-md.svg" alt="Measured on agentmods" 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 | $0.03818 | $0.03818 |
| Opus 5 | $0.01909 | $0.01909 |
| Sonnet 5 | $0.00764 | $0.00764 |
| Haiku 4.5 | $0.00382 | $0.00382 |
Grade A, and why
doc77 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 4d 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 — 259 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
Skill 选择策略
- 简单任务 — 如单行修复、小改动、纯问答、明确单一操作等,直接执行,无需调用 skill
- 复杂任务 — 如新功能开发、重构、bug 诊断、多文件改动、需要设计评审等,自动选择合适的 skill 来处理(如 brainstorming → writing-plans → test-driven-development 等),即使用户没有明确指定 skill
文档语言规范
本项目所有文档遵循以下语言规范:
- 专业名词 / 技术术语:使用 英文 原文(如 Node.js, Express, SQLite, MCP, SDK, API, CLI, SSE, JSON-RPC, TypeScript 等)
- 其余说明文字:使用 中文
术语对照示例
| 英文术语 | 说明 |
|---|---|
| Node.js | 运行时,不翻译 |
| Express | Web 框架,不翻译 |
| SQLite | 数据库,不翻译 |
| MCP (Model Context Protocol) | 协议名称,保留英文 |
| SDK | 软件开发工具包,保留英文缩写 |
| CLI | 命令行接口,保留英文缩写 |
| API | 应用程序接口,保留英文缩写 |
| SSE (Server-Sent Events) | 服务端推送事件,保留英文缩写 |
| JSON-RPC | 协议,保留英文 |
| TypeScript | 编程语言,不翻译 |
| Shadow / Shadow Copy | 影子备份,可保留英文并附中文说明 |
| Pre-flight Check | 飞行前检查 |
| Rollback | 回滚 |
| GC (Garbage Collection) | 垃圾回收 |
| safeMove | 安全移动(函数名保留英文) |
| batch_operations | 批量操作(函数名保留英文) |
| session_id | 会话标识符(字段名保留英文) |
编写原则
- 技术名词首次出现时,可附中文注释说明,后续统一使用英文术语
- 代码、命令、配置项、字段名、表名保持英文原文
- 结构化内容(表格、列表)中的术语使用英文,描述性文字使用中文
- 标题可根据内容性质使用中英混合
公开文档语言规范
项目支持多语言推广。公开文档的语言遵循以下规则:
- 英文(默认/权威版本):
README.md、CHANGELOG.md、package.json#description - 中文翻译:
README.zh-CN.md,通过顶部语言切换 badge 关联 - 内部设计文档(
docs/design/、docs/planning/):保持中文(面向中文开发团队)
隐私与安全规范
严禁将包含隐私数据的文件提交到 Git 仓库。 以下文件一律默认保留在本地,不得 git add、git commit 或 git push:
- 包含 token、API key、secret、password、access_key、private key 等认证凭据的任何文件
- 环境变量文件:
.env、.env.local、.env.development、.env.production等(.env.example模板文件除外) - 证书与密钥:
*.pem、*.key、*.pfx、*.p12、*.cert、*.crt - 任何 OAuth、AWS、GitHub token 等第三方服务凭据
原则
-
默认本地 — 不确定是否包含敏感数据时,默认不提交
-
模板替代 — 如需共享配置结构,提供
.example模板文件(值用占位符) -
提交前隐私检查(强制) — 每次
git commit前,必须对git diff --cached的内容执行隐私泄露扫描,检查是否包含以下敏感信息:类型 检测模式 Token / API Key sk-、ghp_、gho_、xox[bprs]-、AKIA、eyJ(JWT) 等前缀Password / Secret password\s*[:=]、secret\s*[:=]、passwd手机号 1[3-9]\d{9}(中国手机号)邮箱 \S+@\S+\.\S+(非示例/占位邮箱,如[email protected]除外)私钥 -----BEGIN.*PRIVATE KEY-----、-----BEGIN RSA PRIVATE KEY-----内网地址 192\.168\.、10\.\d{1,3}\.\d{1,3}\.\d{1,3}、172\.(1[6-9]|2\d|3[01])\.
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.
- 4d ago First seen · 259 lines · 3,818 tokens per session scan A a654ed5c8235
doc77 CLAUDE.md is an instructions file published in the GitHub repository xyy277/doc77 (8 stars, last pushed 16d ago), licensed MIT. It adds 3,818 tokens to every session, about $0.0191 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 2026-08-31.
Other instructions, from other repositories
abap_wiki AGENTS.md
AGENTS.md instructions for Gixsy95/abap_wiki, covering agents.md - abapwiki operating contract, 1. identity and mission, 2. language, 3. architecture: the two planes and 4. inviolable rules.
gnosis-mcp CLAUDE.md
Instructions for nicholasglazer/gnosis-mcp, covering gnosis mcp -- mcp documentation server, architecture, backend protocol, dependencies and tools.
alcove AGENTS.md
AGENTS.md instructions for epicsagas/alcove, covering claude.md, repository purpose, actual location, structure and karpathy 3-layer architecture.
obsidian-mcp-rs CLAUDE.md
Instructions for MrRefactoring/obsidian-mcp-rs, covering claude.md, commands, build / run, tests and lint gates enforced by ci (.github/workflows/ci.yml).
tela CLAUDE.md
Claude Code instructions for zcag/tela, covering claude.md — tela, what tela is, layout, conventions and run / dev.
accreta CLAUDE.md
Claude Code instructions for francescofioredev/accreta, covering claude.md — accreta, 1. what this project is, 2. origin, and what it means for the code, 3. architecture and 4. conventions.