JavaScript 安全审查
检查 XSS、CSRF、原型污染等安全问题,支持 React/Vue/Angular
Browse reusable Agent Skills, each with a clear purpose and practical guidance.
检查 XSS、CSRF、原型污染等安全问题,支持 React/Vue/Angular
对前端代码进行安全审计,检测 XSS、CSRF 等漏洞。当用户请求代码审查或询问代码安全性时使用。
使用 Vitest 和 Vue Test Utils 的标准模版与断言技巧
Tests Turbo Laravel features in PHPUnit or Pest. Activates when using the InteractsWithTurbo trait; simulating requests with $this->turbo(), $this->fromTurboFrame(), or $this->hotwireNative(); asserting responses with assertTurboStream(), assertNotTurboStream(), assertRedirectRecede(), assertRedirectResume(), or assertRedirectRefresh(); faking broadcasts with TurboStream::fake(), assertBroadcasted(), assertNothingWasBroadcasted(), or assertBroadcastedTimes(); writing feature tests for Turbo Stream responses; or when the user mentions testing Turbo, testing broadcasts, or Turbo test assertions.
Run a comprehensive language-agnostic repository health audit, scored 0-100. Detects language, runs native tooling, never assumes JS/Python.
Diagnostic-only skill for Slurm scheduler and node-daemon issues on Amazon SageMaker HyperPod Slurm clusters. Scope mirrors the HyperPod troubleshooting guide. Invoke when the user reports a Slurm node stuck in down/drain, "Node unexpectedly rebooted" after auto-repair, slurmd not running, jobs stuck PENDING with REASON=Resources while sinfo shows idle nodes, jobs stuck COMPLETING after node replacement, GRES/GPU counts wrong, scontrol ping failing, slurmctld unresponsive, an Action:Reboot/Replace request that did not trigger HyperPod auto-recovery, or auto-resume not restarting a job. Also triggers on "drain before reboot", "diagnose a Slurm node", "investigate stuck jobs."
Audit authentication and session-management code for common issues — weak JWT config, session fixation, password-handling flaws, insecure cookies, broken OAuth flows, and missing auth checks on routes. Use when the user asks to review auth code or when source-aware scanning targets login/session/token handling.
Generate acceptance testing steps after completing a feature or bug fix. Use when the user asks for testing steps, a testing plan, or how to test something.
Use when a Steam, Proton, Wine, Lutris, or Heroic game misbehaves on Linux — crashes at launch, black screen, missing DLSS / Ray Tracing / Path Tracing / DLSS-RR / Frame Generation, anti-cheat failures, shader stutter, controller or HDR or Wayland-fullscreen problems, intro-video failures ("Failed to create SourceReader", MF_E_UNEXPECTED), NVAPI "Unknown function ID" log spam, gamescope dedup black-screens, vkd3d swapchain crashes. Especially before patching or forking DXVK, vkd3d-proton, dxvk-nvapi, gamescope, Wine, or Proton; before reading RE Engine, Unreal, Unity, Source 2, id Tech, or Decima logs; or when a game is mentioned with NVIDIA Blackwell, NixOS, niri, gamescope, proton-cachyos, GE-Proton, or proton-tkg. Triggers on phrases like "won't launch", "RT/PT grayed out", "Proton fix for", "/WineDetectionEnabled", "ProtonDB", "VKD3D_CONFIG", "PROTON_USE_WAYLAND", or any moment the impulse is "let me read the Proton/Wine/vkd3d source" before checking community fixes.