No red flags found in any of the 11 categories — no credential harvesting, no data exfiltration, no curl-pipe-shell installer. Scanned against the SlowMist agent-security taxonomy, refreshed every 8 hours. Full audit →
by oil-oil · Codex Skill · ★ 125
Last updated: · Indexed by AgentSkillsHub · Auto-synced every 8h
🔒 Is oil-frontend safe to install? View the security audit →
一套约束 AI 产品前端实现的 Agent Skill。 任务与对象 · 状态与范围 · 界面与空间 · 组件与代码 为什么需要这个 Skill AI 编写的前端通常可以运行，也可能通过类型检查和构建。实际问题常出现在用户开始操作以后：它可能沿用错误的旧实现，在列表里堆满字段和按钮，把局部请求处理成全页 loading，或者在共享组件外继续补 CSS。改动次数增加以后，界面含义、数据状态和代码归属会逐渐分离。 让 Agent 在设计、实现、重构和评审产品前端时，按照同一套顺序判断问题。 它会处理的五个重点 先确认用户任务和业务对象 AI 容易从已有字段、组件和页面结构继续拼接。Skill 要求先确认用户要完成的任务、页面管理的业务对象、字段所有者和完成结果，再决定使用列表、表格、选择器、表单、弹窗或完整工作区。 项目原有模式本身错误时，不继续复制。 删除无效信息和伪操作 每段文字、图标和动作都必须帮助用户识别、比较、判断、操作或理解结果。没有这些作用的 UUID、内部枚举、、重复流程说明、通用 AI 图标和装饰性强调边会被删除。 每个可见动作都必须产生真实的数据变化、导航或结果。表单只收集当前任务需要的数据，不默认增加草稿、预览、重置、步骤条和重复确认。 保证数据、状态和作用域真实 同一业务数据只保留一个权威来源。选择器不能把当前分页当作全部候选；查询条件、列表、总数和分页必须属于同一次响应；单项请求只锁定对应对象，不阻塞整个页面。 异步操作成功后再关闭弹窗，失败时保留当前对象、用户输入和操作上下文。 让界面结构和代码各自归位 列表负责识别和比较，详情负责完整内容，编辑器负责修改。尺寸由内容和父容器决定，每个页面、弹窗和表格都需要明确自己的滚动范围。 页面负责组合，组件负责自身视觉和状态。组件、Hook、函数、类型、样式和测试按业务归属组织；共享组件的问题在共享层处理，不在页面使用位置反复覆盖样式。 修改源头并完成迁移 Agent 会先检查数据流、父布局、共享组件和全部使用位置，再修改真正产生问题的实现。新实现接管以后，删除旧组件、旧状态、旧类型、旧样式、fallback 和 legacy 逻辑。 验证不会停在类型检查、Lint 或构建通过。Agent 还要检查本次改动涉及的状态、数据范围、视口和共享实现使用位置。 它怎么读取规则 产品前端相关的实现、修改、重构和评审会自动触发这个 Skill。触发后先只读取 SKILL.md 判断任务范围，再按需读取对应参考规则；不相关的任务不会继续加载其他文件。 安装 oil-oil/oil-frontend 安装这个 skill 使用 安装后可以直接提出前端修改
| Stars | 125 |
| Forks | 5 |
| Category | Codex Skill |
| License | MIT |
| Quality Score | 53.0459006289173/100 |
| Last Updated | 2026-09-17 |
| Created | 2026-07-30 |
| Platforms | codex |
| Est. Tokens | ~16k |
Looking for a oil-frontend alternative? If you're comparing oil-frontend with other codex skill tools, these 6 projects are the closest alternatives on Agent Skills Hub — ranked by topic overlap, star count, and community traction.
A modular UX, design, and product optimisation system for turning AI-built interfaces into production-grade ap
A virtual design team for Claude Code, Cursor, Windsurf, Gemini CLI, and Copilot — 26 roles, 62 commands, 15,0
A virtual design studio for Claude Code — 27 slash commands, 13 specialist roles, full design-to-code pipeline
Design intelligence engine for AI coding tools (Claude Code, Cursor, Windsurf). Deterministic anti-AI-slop lin
Design engineering system for AI coding agents — ship UI with craft-level quality. Install as an agent skill.
Consulting firm-style PowerPoint design system for AI agents. 70 layout patterns, flat design, python-pptx. 麦麸
Explore other popular codex skill tools:
oil-frontend is 为产品前端的实现、修改和评审提供中文规范，覆盖界面、交互、数据状态和组件组织。. It is categorized as a Codex Skill with 125 GitHub stars.
You can find installation instructions and usage details in the oil-frontend GitHub repository at github.com/oil-oil/oil-frontend. The project has 125 stars and 5 forks, indicating an active community.
oil-frontend is released under the MIT license, making it free to use and modify according to the license terms.
The top alternatives to oil-frontend on Agent Skills Hub include layr, naksha-studio, design-studio. Each offers a different approach to the same problem space — compare them side-by-side by stars, quality score, and community activity.
Grades come from a rule-based scan built on the SlowMist agent-security taxonomy, covering 11 red-flag categories including credential harvesting, data exfiltration, and curl | sh installers. It is a first-layer scan, not a manual audit — we say so rather than overstate it.
The scale of the problem is documented independently: Liu et al. (2026), in a study of 31,132 agent skills, report that 26.1% contain security vulnerabilities. Our own full-catalog census is published as a citable open dataset.
Sources & who's responsible: