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 martin1847/evolab --skill agent-frontend-standardgit clone --depth 1 https://github.com/martin1847/evolabWrote 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/martin1847/evolab/agent-frontend-standard)<a href="https://agentmods.dev/skills/martin1847/evolab/agent-frontend-standard"><img src="https://agentmods.dev/badge/skills/martin1847/evolab/agent-frontend-standard/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/martin1847/evolab/agent-frontend-standard"><img src="https://agentmods.dev/badge/skills/martin1847/evolab/agent-frontend-standard.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00216 | $0.01839 |
| Opus 5 | $0.00108 | $0.00920 |
| Sonnet 5 | $0.00043 | $0.00368 |
| Haiku 4.5 | $0.00022 | $0.00184 |
Grade A, and why
agent-frontend-standard 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 9d 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 — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Frontend — 前端工程手册
姊妹篇:后端归
agent-backend-standard(hub);Git 归git-workflow-standard;护栏供给回路 canonical =cto-orchestration的 shock-in-the-loop 篇(电在回路/能电不文)。 本手册条目均有 2026 调研证据背书(Anthropic / Figma / Atlassian 等公开来源,二手数据已标注)。 铁律同 hub:"In nearly 100% of cases the prompt will win over the guidelines"(Figma 实证)—— 声明式规则是软的,只有编译器、lint=error 和 CI 是硬的。能电不文。
1. 设计系统即 agent 契约
- 受限组件词表:业务代码只准从少量 vetted DS primitives 组合,禁自由 CSS/裸 HTML 拼装。 机制:"LLM 不擅长一致发明,擅长查找"。
- 薄封装收窄 prop 面:禁 app 直接 import 厂商组件库;wrapper 禁
{...rest}透传——透传=把 厂商全部 API 面暴露给 agent 猜。 - token 语义标注是性价比最高的一招:token 定义处写 role + 反用法("仅 destructive 操作; 禁用于品牌强调")。50 个设计系统审计只有 1 家写了 do-not-use——写了 agent 就稳定选对。
- 三条 lint 硬门(=error,warn 等于没有):
- tailwind
no-arbitrary-value——agent 不知道你的 token 就会吐 hex/px,任意值=样式漂移主通道; no-restricted-imports禁 legacy/厂商直连路径——agent 分不清ui/和legacy/;- 非法 prop 组合用 discriminated union 灭在类型层——"TS 编译器就是设计系统评审员"。
- tailwind
- 机器可查询 > 散文描述:组件 API 给结构化形态(manifest/story/类型),使用规则才写 Markdown。 格式即成本:JSON 组件元数据比散文省 ~80% token(Indeed 1056-prompt 实测,二手)。
2. 类型端到端契约
- 契约在 agent 编辑范围之外:后端契约生成 client,生成物只读——"编不出不存在的 endpoint"。 配对律:契约改动必须触发再生成(pairing 门),生成物手改必须被拦(starter hook ①②已落)。
- typecheck 是 agent 主反馈环:
tsc --noEmit的 file/line/col 五元组比测试堆栈更适合收敛; 每轮都跑(tsgo 已把成本压到可每轮)。 - 封死逃逸口:agent 会用
any/as/@ts-expect-error把红变绿。type-coverage 阈值 + 抑制注释必须人类 override commit 才放行。 - 认知边界:类型错误是 agent 最会修的一类(也正因此 typecheck 绿≠正确);逻辑错误修复率仅 ~45% 且两轮后收益枯竭——正确性预算投给断言强度(mutation/oracle),别再收紧 tsconfig。
3. 视觉验证环
- a11y 树优先、截图兜底(双轨不二选一):snapshot ~200-400 token/确定性/ref 精确;截图 ~3000-5000 token/坐标脆——动画/图表/文案才用截图。axe 类 a11y 检查是少数客观可机验门 (覆盖 ~57%,是门不是证明)。
- 迭代必须设预算上限:视觉环 ≤5 轮、a11y ≤3 轮,超限熔断上报——无退出条件的 agent "要么跑死要么随机停",迭代精化可能只烧成本零改善。
- 多轮视觉反馈会引发功能遗忘(第 N 轮修好、第 N-3 轮坏了)——视觉环必须配回归重检: VRT 基线 diff / story interaction test 每轮全量复跑。这是 VRT 在 agent 环里的真正理由。
- "没有 check 时 looks-done 是唯一停止条件":给 agent 的 UI 任务必须带可跑验证 (story test / screenshot diff / typecheck+lint 组合),同北极星 doctrine。
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.
- 9d ago First seen · 79 lines · 216 tokens per session scan A c1e5a9565c4c
agent-frontend-standard is a skill published in the GitHub repository martin1847/evolab (6 stars, last pushed yesterday), licensed MIT. It adds 216 tokens to every session and 1,839 once invoked, about $0.0011 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
screen-reader-testing
Test web applications with screen readers including VoiceOver, NVDA, and JAWS. Use when validating screen reader compatibility, debugging accessibility issues, or ensuring assistive technology support.
test-quality
Write high-quality JUnit 5 tests with AssertJ assertions. Use when user says "add tests", "write tests", "improve test coverage", or when reviewing/creating test classes for Java code.
react-web
React web development with hooks, React Query, Zustand.
app-user-story-qa
End-to-end app feature inventory and user-story testing workflow with a canonical tracker. Use when the user asks to audit every feature, derive expected behavior from code, test user journeys, or explicitly fix and retest documented UX or logistical defects.
ln-23-test-suite-auditor
Audits existing tests for meaningful coverage, trustworthy oracles, and maintenance value. Not for test implementation or a single delivery review.
ln-21-documentation-auditor
Audits documentation and comments for trustworthy claims, coverage, and discoverability. Not for code, test, or architecture audits.