菜单
Generate Crucible end-to-end Playwright specs by dispatching the local playwright-test-generator agent. Use when asked to implement one or more scenarios from a Crucible app's test plan and verify them against the running stack.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/cmu-sei/crucible-tests --skill generate-crucible-tests命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Diagnose and repair failing Crucible end-to-end Playwright tests by dispatching the local playwright-test-healer agent. Use when a Crucible test fails, flakes, has stale selectors, or needs browser-based root-cause investigation and a focused fix.
Create or expand a Crucible end-to-end Playwright test plan by dispatching the local playwright-test-planner agent. Use when asked to explore a running Crucible app, document browser test scenarios, or update an app's test-plan markdown.
基于 SOC 职业分类
正在显示 SKILL.md
| name | generate-crucible-tests |
| description | Generate Crucible end-to-end Playwright specs by dispatching the local playwright-test-generator agent. Use when asked to implement one or more scenarios from a Crucible app's test plan and verify them against the running stack. |
Dispatch the repository's playwright-test-generator agent. Keep browser
execution and spec generation in that agent so it can use its generator tools.
multi_agent_v1.spawn_agent with
agent_type: "playwright-test-generator" and that exact scope. Tell it to
work in the current repository, update the relevant spec files, and run the
narrowest useful verification.multi_agent_v1.wait_agent for completion, then report changed files,
verification results, and any blockers.