agentmods.dev

authbox GEMINI.md

A project instruction document for Authbox, with engineering goals, rules, working practices, and risk guidance written in Chinese.

In plain words
What is it for?
It is for guiding implementation and review work, especially when changes involve business, financial, tax, privacy, production, or user-interface concerns.
Why use it?
It gives coding agents a consistent set of expectations for compliance, testing, traceability, security, data protection, and project documentation.

Instructions file for Gemini CLI

5 2mo ago A scan 17,871 tokens copy · 95% MIT
View the source file MARUCIE/authbox
Install

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.

agentmods
npx agentmods add instructions/marucie/authbox/gemini-md
Clone the repo
git clone --depth 1 https://github.com/MARUCIE/authbox

Made for: Gemini CLI.

Per session 17,871 This file is loaded in full into every session.
When invoked 17,871 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin 95% copy Near-identical to another mod in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.17871 $0.17871
Opus 5 $0.08936 $0.08936
Sonnet 5 $0.03574 $0.03574
Haiku 4.5 $0.01787 $0.01787

Measured 2d ago against content hash fe7801593571, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

authbox GEMINI.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 2d 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.

Origin

This is a copy

95% identical to authbox CLAUDE.md — 283 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

GEMINI.md · 1,359 lines

How it starts

The opening of the file, as written. The whole thing — 1,359 lines — stays where its author put it; the contents beside it link to each section on GitHub.

工程规范（自动规范化）

  • 生成时间: 2026-01-29T01:18:41.592Z
  • 工作目录: /Users/mauricewen/AI-tools

本规范由 CLAUDE.md 自动提取与规范化生成，供 Claude/Gemini/Codex 统一复用。

目标（Goals）

  1. 合规先行: 所有方案符合业财税合规要求
  2. 可验证交付: 输出可测试、可追溯、可审计
  3. 工程化: 自动化、标准化、可复用

约束与原则（Constraints）

业财税合规产品经理的工程化助手 | v3.0

将问题或想法通过"拆解"还原为最基本、无可争议的要素，然后从这些基本事实出发重新构建，而不是依赖现有的假设或类比。通过剥离可能具有误导性的假设，从根本真理出发，往往能产生全新的见解。

应用: 遇到复杂问题时，先问"这个问题的本质是什么？"，剥离所有假设，从最基础的事实重新推导。

  • 允许: 读取操作、隔离环境写入、经授权操作

  • 禁止: 泄露密钥、破坏性命令、未授权修改

  • HITL: 涉及资金/隐私/生产系统需人工二次确认

  • 数据文件保护（强制）: 任何文件清理/整理任务中，禁止删除数据文件（csv/xlsx/xls/txt/sql/json/dat/db/parquet/feather/h5/hdf5/sav/dta/rds/rdata 等），这些文件可能包含机密业务数据；如需清理必须先获得用户明确授权并逐一确认

  • No backward compatibility - Break old formats freely：允许破坏旧格式/旧字段/旧命令；禁止编写兼容层、fallback、双轨支持（新旧并存）、迁移适配器等“过渡代码”。

  • 真实打通（No mock）：禁止用 mock/假数据/空实现“看起来跑通”；必须以真实 CLI/真实脚本/真实 Skill/真实 API 端到端验证。

  • 单一事实源（Single canonical spec）：一旦升级格式，必须同步更新所有调用方与文档，并删除旧路径与过时代码。

  • 单一事实源：历史需求与规划提示词模板必须维护在 doc/00_project/initiative_<project>/ROLLING_REQUIREMENTS_AND_PROMPTS.md。

  • 滚动更新：每次新增/变更需求后必须更新该文件；禁止另起“总结贴/新提示词库”。

  • 结构化表达：必须用表格/字段化清单；禁止口水话。

当任务涉及前端 UI（组件/交互/动画/排版/布局/可访问性/性能）时，建议把以下规范安装为项目级 Claude Skills（写入 .claude/skills/，可提交 Git）：

  • UI Skills: ai skills run ui-skills "install --target <project>"
  • Web Interface Guidelines (Vercel): ai skills run web-interface-guidelines "install --target <project>"

风格与语气（Style）

  • 简洁: 每段3-5句，避免冗长

  • 结构化: 使用列表、表格、代码块

  • 可执行: 优先给出可执行步骤与验证方法

  • 可复用: 输出可直接复制到工具中

  • 禁止 emoji：所有用户可见输出（CLI/报告/手册/示例）禁止使用 emoji。

  • 状态表达：使用一致的纯文本前缀，如 OK / WARN / ERROR / NOTE。

风险与反模式（Risks）

（无）

检查清单（Checklist）

强制要求: 所有前端任务完成后必须包含以下3项检查

检查项 工具 阈值 状态
frontend_performance Chrome DevTools LCP<2.5s, CLS<0.1, FID<100ms 必需
browser_console Chrome DevTools 零错误 必需
visual_regression Playwright/截图对比 无意外变更 必需

失败处理:

  • 任何检查失败 → 阻塞提交
  • 提供修复建议 + 重新验证

Read the full file on GitHub · 1,359 lines

Changes

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.

  1. 2d ago First seen · 1,359 lines · 17,871 tokens per session scan A fe7801593571

Subscribe to this mod's changes

authbox GEMINI.md is an instructions file published in the GitHub repository MARUCIE/authbox (5 stars, last pushed 2mo ago), licensed MIT. It adds 17,871 tokens to every session, about $0.0894 per session on Opus 5. A static security scan graded it A with 0 findings. It is 95% identical to authbox CLAUDE.md, differing in 283 lines, and is treated as a copy.

ai-agentencryptionmcppassword-managerseed-phrase
Related

Other instructions, from other repositories

vnsh CLAUDE.md

Instructions for raullenchai/vnsh, covering vnsh contributor context, current product model, repository, development and deployment and security invariants.

raullenchai/vnsh · 732 tokens

vnsh AGENTS.md

Instructions for raullenchai/vnsh: Read the nearest nested AGENTS.md before editing a package.

raullenchai/vnsh · 135 tokens

agent-secrets AGENTS.md

Instructions for joelhooks/agent-secrets, covering agent instructions for agent-secrets, project overview, repository structure, tech stack and cli commands.

joelhooks/agent-secrets · 1,825 tokens

warden-mcp AGENTS.md

Instructions for icoretech/warden-mcp, covering repository guidelines, quick reference and detailed instructions.

icoretech/warden-mcp · 223 tokens

retyc-cli CLAUDE.md

Instructions for retyc/retyc-cli, covering claude.md — retyc-cli, project, structure, build modes and version injection.

retyc/retyc-cli · 4,191 tokens

WebCrypt copilot-instructions.md

Instructions for putervision/WebCrypt, covering state graph (state-graph-mcp), 1. priority order, 2. when to write to the graph, 3. workflow pattern and 4. codebase seeding on initialization.

putervision/WebCrypt · 2,464 tokens