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 BackToCimaCoppi/Praxis --skill test-standardsgit clone --depth 1 https://github.com/BackToCimaCoppi/PraxisWrote 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/backtocimacoppi/praxis/test-standards)<a href="https://agentmods.dev/skills/backtocimacoppi/praxis/test-standards"><img src="https://agentmods.dev/badge/skills/backtocimacoppi/praxis/test-standards/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/backtocimacoppi/praxis/test-standards"><img src="https://agentmods.dev/badge/skills/backtocimacoppi/praxis/test-standards.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 | $0.00104 | $0.02325 |
| Opus 5.5 | $0.00042 | $0.00930 |
| Sonnet 5.5 | $0.00021 | $0.00465 |
| Haiku 4.5 | $0.00010 | $0.00232 |
Grade A, and why
test-standards 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 1mo 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 — 126 lines — stays where its author put it; the contents beside it link to each section on GitHub.
测试规范层
本 skill 只回答：这次改动必须测什么、测到什么深度、什么情况不能交付。不要在本层写具体命令、账号、端口、脚本写法或项目工具细节；执行方式交给 test-execution-router 与项目执行 skill。
0. 第一原则
在纯 AI 开发 loop 中，测试是最后一道质量闸，不是可选收尾。没有测试矩阵、没有自动化执行证据、没有无法自动化的明确理由，不允许交付。
默认规则：
- 能由 AI 自动化的测试必须自动化。
- AI 可控制浏览器/开发者工具、读取 DOM/data、截图并用 VLM 判读时，仍属于自动化；“有 GUI”“需要看图”“工具是交互式”都不是手工理由。
- 只有 AI 无法操作的真实设备、原生客户端对象、支付/OAuth/原生授权等物理边界，才允许转人工手工验证。
- 手工验证必须有物理边界证明、runbook、客观观测点、回传材料和写明责任角色的计划内人工里程碑。
- 跳过任一测试类型时，必须说明本次改动为什么不命中，而不是写“暂不需要”。
- 每次测试矩阵对下列八面恰好各有一行；不命中写可核 N/A 理由，禁止省略行。
1. 必选测试面
每次改动都先判断是否命中下列测试面，并形成测试矩阵：
| 测试面 | 何时必须纳入 | 最低要求 |
|---|---|---|
| 单元测试 | 业务规则、状态分支、算法、解析、组装、校验逻辑变化 | 覆盖正常、边界、异常或回归路径；不能只测 happy path |
| 接口测试 | 对外接口、鉴权、错误码、状态流、请求/响应字段变化 | 断言字段语义、错误语义、状态变化；不能只看成功码 |
| 数据库数据预置 | 需要稳定前置状态、跨步骤业务链路或复杂终态校验 | 幂等、可清理、命名空间隔离；优先通过业务接口造数 |
| 数据库数据校验 | 写链路、状态流转、表结构、索引/唯一约束、异步重算变化 | 校验终态、不变量、幂等和约束；不能只看有记录 |
| 网页 UI 自动化 | Web/admin 页面、表单、列表、弹窗、配置页、可视状态变化 | AI 自动操作关键路径，硬断言 DOM/数据，必要时自动截图+VLM 判读 |
| 小程序 UI 自动化 | 小程序页面、组件、状态管理、C 端路径变化 | AI/自动化工具操作关键路径，硬断言元素/data，canvas/复杂视觉自动截图判读 |
| 小程序视觉验收 | 页面布局、图表、canvas 合成图、视觉规格变化 | 自动产出截图并由 VLM/视觉模型判读规格差异；不能只证明截图存在 |
| 手工验证 | AI 无法操作的真实设备、原生客户端对象、支付/OAuth/原生授权 | 仅作物理边界例外；物理边界证明 + runbook + 观测点 + 回传材料 + 人工里程碑及责任角色缺一不可 |
2. 改动范围到测试矩阵
按改动范围选择测试组合：
| 改动范围 | 必跑 | 常见附加 |
|---|---|---|
| 纯后端内部规则 | 单元测试 | 若影响外部行为，加接口测试；若落库，加 DB 校验 |
| 对外接口 | 接口测试、相关单元测试 | 若写数据，加 DB 预置/校验；若被 UI 消费，加对应 UI 自动化 |
| 数据库结构或写链路 | DB 预置、DB 校验、相关单元/接口测试 | 若影响展示，加 UI 自动化或视觉验收 |
| 后台网页 | Web UI 自动化、接口测试 | 配置写入必须加 DB 校验；视觉变化加截图判读 |
| 小程序业务页 | 小程序 UI 自动化、接口测试 | 视觉/图表/canvas 变化加视觉验收 |
| 小程序视觉/合成图 | 小程序 UI 自动化、小程序视觉验收 | 若素材/配置来自后端，加接口测试与 DB 校验 |
| 鉴权、支付、用户数据删除、业务 ID、核心算法 | 相关测试全纳入，金标准/回归必跑 | 自动升级为高风险质量闸；测试通过不等于免除人工审查，但人工审查发生在设计/章程拍板与终点验收，不构成执行中的停机闸门 |
| 纯文案或非行为样式微调 | 视影响面最小化 | 仍需视觉或 UI 验证；不得默认“无需测试” |
3. 数据策略
根据业务真实性和风险选择数据策略：
| 策略 | 适用 | 要求 |
|---|---|---|
| DB 预置 -> 测试 -> DB 校验 | 需要稳定构造复杂状态、异常状态或历史数据 | 前置必须幂等、可清理；不得 broad cleanup；业务 ID 需合法来源 |
| 接口造数 -> 测试 -> DB 校验 | 需要验证真实业务写链路或 ID/状态由系统生成 | 优先使用；造数接口本身也要记录为前置链路 |
| 复用测试数据 -> 测试 -> DB 校验 | 数据昂贵或依赖外部平台 | 必须证明数据稳定、不会污染他人测试 |
| 人工前置 -> AI 自动校验 | 真机/外部平台授权等 AI 无法完成的前置 | 人工只做绕不过去的动作；后续校验仍由 AI 执行 |
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 1mo ago First seen · 126 lines · 104 tokens per session scan A 0f8f69ce3074
test-standards is a skill published in the GitHub repository BackToCimaCoppi/Praxis (6 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 104 tokens to every session and 2,325 once invoked, about $0.0004 per session on Opus 5.5. 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
emulate-seed
Generate emulate seed configs for stateful API emulation. Wraps Vercel's emulate tool for GitHub, Vercel, Google OAuth, Slack, Apple Auth, Microsoft Entra, AWS, Okta, Clerk, Resend, Stripe, and MongoDB Atlas APIs — full state machines, not mocks. Use when setting up test environments, CI pipelines, integration tests…
architecture-patterns
Architecture validation and patterns for clean architecture, backend structure enforcement, project structure validation, test standards, and context-aware sizing. Use when designing system boundaries, enforcing layered architecture, validating project structure, defining test standards, or choosing the right…
error-analysis
Evals-first error analysis for LLM apps: clusters real Langfuse or JSONL traces into a human-confirmed failure taxonomy with counts, then recommends binary pass/fail evals for recurring named modes. Use to learn what to measure before writing evals. Not for CI failures.
testing-e2e
End-to-end testing patterns with Playwright — page objects, AI agent testing, visual regression, accessibility testing with axe-core, and CI integration. Use when writing E2E tests, setting up Playwright, implementing visual regression, or testing accessibility.
testing-integration
Integration and contract testing patterns — API endpoint tests, component integration, database testing, Pact contract verification, property-based testing, and Zod schema validation. Use when testing API boundaries, verifying contracts, or validating cross-service integration.
testing-llm
LLM and AI testing patterns — mock responses, evaluation with DeepEval/RAGAS, structured output validation, and agentic test patterns (generator, healer, planner). Use when testing AI features, validating LLM outputs, or building evaluation pipelines.