Watch an open PR and fix its CI failures and review feedback until it's clean. Use standalone for a PR you want monitored, or as the babysit step of /ship, which continues through the merge.

The GitHub issues tracker for refactor findings: issues labeled refactor-findings in the fragility repo, with the plan in the body. Read when fragility-common/config.json sets tracker to github and you need to know how findings map onto issues, or when changing that mapping.

Diagnose intermittent VS Code Electron smoke-test failures from the Azure DevOps Flaky Smoke Tests pipeline (definition 700). Covers finding failed iterations, downloading task logs and platform artifacts with Azure CLI, correlating cumulative runner logs, tracing the introducing commit, and queueing focused validation runs.

This skill should be used when the user says "solve problem", "solve this problem", or invokes /solve-problem. Applies a methodical, evidence-first approach to diagnosing a problem - understand intent, observe the problem first-hand, form multiple hypotheses, gather discriminating evidence to eliminate them, propose an evidenced solution, and verify the fix actually resolves the original symptom. Do NOT use for general how-to questions, explanations, or building new features.

This skill should be used when the user asks to "create a skill", "write a skill", "build a skill", or wants to add new capabilities to Claude Code. Use when developing SKILL.md files, organizing skill content, or improving existing skills. Do NOT use for plugin development, hook creation, agent creation, or slash command creation — those have dedicated skills.

Prepare a plain-language PR review briefing from a GitHub pull request URL or number. Use when the user says "help me review this PR", "summarize this PR for review", "what does this PR do", or wants the reason, systems involved, changes, risks, CI state, and questions needed before reviewing a pull request.

短篇网文扫榜。分析知乎盐言、七猫、黑岩、点众等平台热门短篇数据，捕捉风口题材。触发方式：/story-short-scan、/短篇扫榜、「短篇什么火」「知乎故事排行」。

Vercel deployment and CI/CD expert guidance. Use when deploying, promoting, rolling back, inspecting deployments, building with --prebuilt, or configuring CI workflow files for Vercel.

Diagnose measured MSBuild evaluation cost before targets run. USE FOR: evaluation-only repeats under global-property variants, expensive node_modules/.git/bin/obj globs, import depth >20, /pp >10K lines, file-I/O property functions, five-phase evaluation. DO NOT USE FOR: duplicate executions (find-repeated-work-within-one-build), execution bottlenecks (build-perf-diagnostics), incremental rebuilds (incremental-build), non-MSBuild systems.

Use when the user explicitly asks for a CI check or to push their branch — e.g. "ci readiness", "check ci", "pre-push check", "ready for CI", "ci check", "ready to push", "push my changes", "push the branch", "let's push". Catches common CI failures before pushing — formatting, stale API reports, missing changesets, policy violations.

Install and register the model-free AutoRAG Lite MCP server, configure approved roots and datasources, build indexes, and verify MCP discovery and search. Use when autorag-mcp is missing, MCP connection fails, indexes are stale, or the user wants document search without configuring a search model.

Diagnose and fix retention problems using behavior design (B=MAP). Use when the user mentions "users drop off", "activation rate", "onboarding friction", "retention metrics", "why users dont complete", "churn analysis", "user activation", or "aha moment". Also trigger when analyzing cohort retention curves, designing activation milestones, reducing time-to-value for new users, or investigating why users stop after their first session. Covers the Ability Chain, prompt design, and tiny behaviors that compound. For habit loops and variable rewards, see hooked-ux. For intrinsic motivation, see drive-motivation.

Audit websites and landing pages for conversion issues and design evidence-based A/B tests. Use when the user mentions "landing page isnt converting", "conversion rate", "A/B test", "why visitors leave", "objection handling", "bounce rate", "split testing", or "conversion funnel". Also trigger when diagnosing why signups are low, designing experiment hypotheses, or auditing checkout flows for friction points. Covers funnel mapping, persuasion assets, and objection/counter-objection frameworks. For overall marketing strategy, see one-page-marketing. For usability issues, see ux-heuristics.

Apply the six principles of ethical persuasion (reciprocity, commitment, social proof, authority, liking, scarcity) to product design, copy, and sales. Use when the user mentions "social proof", "persuasive copy", "why users dont convert", "ethical persuasion", "reciprocity", "scarcity tactics", or "commitment and consistency". Also trigger when designing testimonial sections, crafting urgency messaging, or improving trust signals on landing pages. For deal negotiation tactics, see negotiation. For viral word-of-mouth, see contagious.

SwiftUI architecture patterns, state management with @Observable, view composition, navigation, performance optimization, and modern iOS/macOS UI best practices. Use when building or reviewing SwiftUI views, @Observable state, navigation, or render performance.

Visualize whether skills, rules, and agent definitions are actually followed — auto-generates scenarios at 3 prompt strictness levels, runs agents, classifies behavioral sequences, and reports compliance rates with full tool call timelines. Use when checking whether agents actually follow the skills, rules, and definitions they were given, rather than assuming they do.

Nuxt 4 app patterns for hydration safety, performance, route rules, lazy loading, and SSR-safe data fetching with useFetch and useAsyncData. Use when building or reviewing a Nuxt 4 app, or debugging hydration mismatches and SSR-safe data fetching.

Manages Android AppWidget state updates, background syncs, and RemoteViews bitmap memory limits with Jetpack Glance (androidx.glance:glance-appwidget) and AppWidgetProvider receivers on Android 17 (API 37). Use this skill when configuring Android 17 (API level 37) compileSdk/targetSdk and AndroidManifest.xml widget receivers, refreshing frozen or stale Glance widget UI after CoroutineWorker DataStore preference updates (updateAppWidgetState), downsampling images within the WindowMetrics memory limit for SizeMode.Responsive, SizeMode.Exact, or LazyColumn, sharing content:// URIs with FileProvider, or fixing ActionCallback NoSuchMethodException and AppWidgetProvider ForegroundServiceStartNotAllowedException crashes with WorkManager. Don't use for Glance layout templates (suggest the glance-canonical-layouts skill) or in-app Compose screens.

Helps developers integrate, debug, and resolve Play Engage SDK implementation issues. Use when adding Engage SDK support, generating publishing code, mapping data classes to entities, or fixing SDK-related errors.

Set up or use the OpenSpace component of HKUDS AgentPilot for retrieving and retaining agent skills. Use when the user asks for OpenSpace skill management or its MCP connection; ordinary coding improvement uses agentpilot.