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/lethechen/init-claude-md/agents-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/instructions/lethechen/init-claude-md/agents-md)<a href="https://agentmods.dev/instructions/lethechen/init-claude-md/agents-md"><img src="https://agentmods.dev/badge/instructions/lethechen/init-claude-md/agents-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 | $0.00868 | $0.00868 |
| Opus 5 | $0.00434 | $0.00434 |
| Sonnet 5 | $0.00174 | $0.00174 |
| Haiku 4.5 | $0.00087 | $0.00087 |
Grade A, and why
init-claude-md AGENTS.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 3d 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.
Copies of this mod
1 near-identical copy found in the catalogue:
- init-claude-md CLAUDE.md — 91% identical, 3 lines differ
How it starts
The opening of the file, as written. The whole thing — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Session Maintenance Protocol
开始时:读 TODO.md,复述当前 🔴 和 🟡 段;扫 references/ 是否与 SKILL.md 描述一致。
结束前(最后一条用户消息之后、停止前必做):
- 回看本次会话实际做了什么 —— 改动的文件、发现的问题、用户新增的约束
- 更新
TODO.md(决策表见references/session-maintenance-protocol.md) - 更新本文件的规则 —— 本次会话暴露的新约定加到对应段
- 刷新日期戳
> 最后更新:YYYY-MM-DD
触发条件:≥1 个仓库文件被改 · 用户新增约束 · 会话 > 10 分钟
Core rules (from Karpathy)
- Think Before Coding — surface assumptions, push back when warranted
- Simplicity First — minimum code that solves the problem
- Surgical Changes — touch only what you must
- Goal-Driven Execution — define success criteria, loop until verified
Language
- Communication:
zh(用户偏好) - Code comments:
zh - Commit messages:
en(Conventional Commits 必须英文) - Rule files:
zh
Project conventions
- 纯 Markdown skill:没有代码 / 包管理器 / 测试运行器
- 编辑
SKILL.md时必须同步更新references/和examples/(决策表、模板、检测矩阵三者锁步) - 所有文件统一用 UTF-8 no BOM(PowerShell
Set-Content -Encoding UTF8默认带 BOM,禁止用) references/session-maintenance-protocol.md是 single source of truth,根文件里的短版必须与之保持一致
Workflow
- Branch:
codex/<scope>-<slug>(solo 维护时直接用 main,feature branch 可选) - Commit: Conventional Commits(
feat:,fix:,chore:,docs:,refactor:) - Solo 维护者:直接
git push origin main,commit log 即 audit trail - TODO.md:每次会话结束时按决策表更新
Verification
git diff --stat HEAD 确认改动符合预期;改 SKILL.md 后用 Get-Content -Encoding UTF8 检查没有 BOM。
Things to avoid
- 不要在 skill 主体加 TypeScript / 任何编译型语言
- 不要加 CI 依赖(这个 skill 没必要要求 Node 18+ 才能用)
- 不要改
SKILL.md时漏改references/(决策表 / 模板 / 检测矩阵三者必须锁步) - 不要写带 BOM 的 UTF-8 文件(PowerShell 默认行为)
- 不要从
examples/删文件而不补 README 的目录树 - 不要把
description字段里的中文触发词写错(之前是涓?乱码,每个 agent 看到的触发词是错的) - 不要用 PowerShell Get-Content | Set-Content 改 Markdown(CRLF 会丢成 LF-only、\a \b \v 会被 PS 解释成控制字符)。用 [System.IO.File]::ReadAllBytes/WriteAllBytes 字节级操作
- 不要在 skill 里加 hook / cron / 外部触发器(软自动才是设计核心;外部触发器破坏跨 agent 兼容)
- 不要去掉根文件顶部的
## Session Maintenance Protocol块(agent 失去结束前自检清单)
最后更新:2026-07-03
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.
- 3d ago First seen · 54 lines · 868 tokens per session scan A d4d46f64baed
init-claude-md AGENTS.md is an instructions file published in the GitHub repository LetheChen/init-claude-md (1 stars, last pushed 1mo ago), licensed MIT. It adds 868 tokens to every session, about $0.0043 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
AutoVideo-Agent AGENTS.md
Instructions for wangxin6x/AutoVideo-Agent, covering agents.md, project rules and development.
codex-command-center AGENTS.md
AGENTS.md instructions for sevensir888/codex-command-center, a project described as: A local desktop control center for organizing Codex projects, engineering tasks, sessions, Git changes, Skills, and MCP configuration.
GameStudio-CodexKIT AGENTS.md
AGENTS.md instructions for hoatv2211/GameStudio-CodexKIT, covering gamestudio-codexkit operating contract, active sources of truth, primary distribution, repository-local maintenance and evidence.
zhijuan-codex-agency-chief-of-staf AGENTS.md
Instructions for papperrollinggery/zhijuan-codex-agency-chief-of-staf: 当当前 Git 根目录是本仓库时,默认任务是维护 Agency Chief of Staff Skill 的源码,而不是运行已安装的 Skill。.
arxiv-wiki AGENTS.md
AGENTS.md instructions for synabreu/arxiv-wiki, covering agents.md, project goal and rules.
ultra-codex-agents.md AGENTS.md
AGENTS.md instructions for keithhegit/ultra-codex-agents.md, covering global coding discipline, think before coding, simplicity first, surgical changes and goal-driven execution.