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 LetheChen/init-claude-md --skill init-claude-mdgit clone --depth 1 https://github.com/LetheChen/init-claude-mdWrote 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/lethechen/init-claude-md/init-claude-md)<a href="https://agentmods.dev/skills/lethechen/init-claude-md/init-claude-md"><img src="https://agentmods.dev/badge/skills/lethechen/init-claude-md/init-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 5.1 | $0.00137 | $0.02778 |
| Opus 5 | $0.00068 | $0.01389 |
| Sonnet 5 | $0.00027 | $0.00556 |
| Haiku 4.5 | $0.00014 | $0.00278 |
Grade A, and why
init-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 8d 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 — 206 lines — stays where its author put it; the contents beside it link to each section on GitHub.
生成 CLAUDE.md / AGENTS.md + TODO.md
为项目生成项目专属的 AI 编码规则文件（CLAUDE.md / AGENTS.md）+ 活的 TODO.md 任务清单。
两个核心模式：
- HumanLayer 极简模式 — 根文件 ≤60 行，不写目录树 / 架构概览 / 不可验证的废话
- Karpathy 4 条准则 — Think Before Coding / Simplicity First / Surgical Changes / Goal-Driven Execution
关键机制：把 ## Session Maintenance Protocol 块嵌入到根文件顶部，未来的 agent 加载根文件时自动按决策表自检更新 TODO.md 和规则——软自动，零依赖（详见末尾「更新机制」）。
工作流
Step 0：平台与目标文件检测
按以下信号判定输出文件名（多个信号取并集）：
| 信号 | 偏好 |
|---|---|
| 存在 .claude/ 目录 | Claude Code（写 CLAUDE.md） |
| 存在 .codex/ 目录 | Codex（写 AGENTS.md） |
| 已 commit 的 CLAUDE.md | 保留，写 CLAUDE.md |
| 已 commit 的 AGENTS.md | 保留，写 AGENTS.md |
| 用户显式说 Codex / Claude Code / 两个都要 | 用户说了算 |
| 都检测不到 | 默认双写（最安全） |
不要在没有信号时强行猜。
Step 1：项目扫描
并行读取（Codex 用 shell_command，Claude Code 用 Read）：
- Manifest：package.json / pyproject.toml / go.mod / Cargo.toml / pom.xml → 语言、框架、scripts
- README：前 50 行 → 项目意图 + 主语言
- 现有规则文件：CLAUDE.md / AGENTS.md（合并策略见 Step 3）
- 现有 TODO.md（迁移但不覆盖）
- 代码风格：.eslintrc* / biome.json / .prettierrc*
- 任务运行器：Makefile / justfile / package.json scripts
- CI：.github/workflows/
- Monorepo 检测：顶层是否含 apps/ packages/ services/ modules/
- TODO 注释：扫一次 TODO / FIXME / HACK（为 TODO.md 提供初始项）
空项目（无代码）→ 停下来问用户，不要编造。
Step 2：决定文件布局
是否有清晰技术栈？ ├── 否 → 问用户，不要生成 └── 是 │ 是否 monorepo（apps/ packages/ services/ 多个子包）？ ├── 是 → Layout C └── 否 │ 源代码 < 10 且无 DB / 无 auth？ ├── 是 → Layout A（仅根文件） └── 否 → Layout B（根文件 + 规则目录，默认）
- Layout A — 仅根文件 CLAUDE.md / AGENTS.md
- Layout B — 根文件 + .claude/rules/*.md（默认）
- Layout C — Monorepo 根文件 + 每个子包一份规则
Step 3：合并策略（如目标文件已存在）
| 段 | 处理 |
|---|---|
| 缺失段 | 添加并填入检测结果 |
| 段内容是占位符（<...> / 空） | 填充 |
| 段内容是真实文本 | 保留，不覆盖 |
| Things to avoid 段 | 追加到末尾，不删现有条目 |
| Verification 命令 | 仅在原命令跑不通时更新；保留用户已写好的命令 |
| Workflow 分支/提交规范 | 保留用户约定，不擅自改成「标准」约定 |
| Session Maintenance Protocol 段 | 重新生成（skill 自身维护的，不属于用户手写） |
| Language 段 | 重新生成（4 字段必填） |
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.
- 8d ago First seen · 206 lines · 137 tokens per session scan A 4ec4a175ec69
init-claude-md is a skill published in the GitHub repository LetheChen/init-claude-md (1 stars, last pushed 2mo ago), licensed MIT. It adds 137 tokens to every session and 2,778 once invoked, about $0.0007 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 skills, from other repositories
boostprompt
Use quando uma necessidade de negócio ou técnica precisar de discovery estruturado em pt-BR e de um prompt de implementação autocontido.
image-first-frontend
Generate a polished frontend reference screenshot through a configurable GPT-image-2-compatible generations endpoint, save and show versioned previews inside the project, revise selected previews through a configurable /images/edits endpoint, and reproduce only the user-approved design as functional frontend code with…
ai-cli-bootstrap-skill
Build and run tiny AI-only CLI tools to unblock tasks that depend on data or actions the agent cannot natively access. Use when the agent is asked to read, collect, convert, or analyze content from external platforms or APIs such as X posts, YouTube transcripts, Reddit threads, web pages, feeds, local databases, or…
tastemaker
Generate genuinely beautiful, on-brand UI instead of generic "AI slop" — use whenever the user asks to build, design, style, or improve a UI, landing page, dashboard, app screen, or component, whenever a PRD/spec needs a design pass before implementation, whenever the user pastes reference images/Pinterest/Dribbble…
codex-autoresearch
Run autonomous, measurable experiments in a Git repository: change one hypothesis, verify a numeric metric, keep improvements, and revert failures. Use when the user wants Codex to keep iterating toward a numeric target in the foreground or as a detached background run. Do not use for ordinary one-shot coding…
fde
Keeps the engagement record for client work. Use when they name a client or stakeholder. Use when they debrief a meeting or paste notes. Use when they ask what was agreed. Use when they run a POC, change the client's codebase, prove it on their staging, go live, or need evals before a model acts. Use when they prep a…