发现 Skills

Uses Chrome DevTools via MCP for efficient debugging, troubleshooting and browser automation. Use when debugging web pages, automating browser interactions, analyzing performance, or inspecting network requests. This skill does not apply to `--slim` mode (MCP configuration).

How to properly create Claude Code skills - teaches global vs project-local, naming conventions, and proper file structure

Implements asynchronous programming patterns using Task, ValueTask, and ConfigureAwait in .NET. Use when building non-blocking I/O operations or improving application responsiveness.

视频创作技能。图片+音频合成视频，支持TTS配音、淡入淡出转场、字幕、片尾、BGM。当用户提到「生成视频」「做视频」「教学视频」「图文转视频」「做视频号」「配音视频」「图文结合视频」「古诗视频」「故事视频」时触发。内含生图→配音→合成全流程，无需单独调用image-service。

This skill should be used when the user asks to "analyze network traffic with Wireshark", "capture packets for troubleshooting", "filter PCAP files", "follow TCP/UDP streams", "dete...

Manage Alibaba Cloud Simple Application Server (SWAS OpenAPI 2020-06-01) resources end-to-end. Use for querying instances, starting/stopping/rebooting, executing commands (cloud assistant), managing disks/snapshots/images, firewall rules/templates, key pairs, tags, monitoring, and lightweight database operations.

Control Philips Hue lights/scenes via the OpenHue CLI.

Analyze data efficiently

Generate and run E2E tests using browser-use AI automation. This skill should be used when creating automated browser tests, testing authenticated flows, generating test scripts from natural language, or validating user journeys with AI-powered browser control. Handles credentials securely via .env.test with domain-prefixed variables.

Launch a real visible browser window (headed mode) instead of headless. Use browser_use with the headed parameter to open, snapshot, click, etc. Ideal for demos, debugging, or when the user wants to see the page.

Run a retrospective on recent work. Analyzes what went well, what didn't, and captures improvements for future sessions.

Spawn an agent team for parallel work — review, research, implementation, or debugging. Configures teammates with appropriate roles and coordinates their work.

Capture solved problems as categorized documentation with YAML frontmatter for fast lookup

Enforce PII table separation, envelope encryption, and KEK rotation for Prisma + Postgres.

Interact with Moltbook (the AI agent social network) — post, browse, comment, vote, search — via shell curl.

Fulcrum: cppbtc-team Electrum-protocol indexer in C++. Faster than electrs for high-volume queries. Serves Electrum clients + LN nodes. USE WHEN: deploying high-load Electrum server, comparing with electrs.

Jetpack Compose for native Android UI. Covers @Composable functions, state, side effects, ViewModel + Compose, Hilt DI, Navigation Compose, Material 3 with Dynamic Color (Material You), Compose for Wear OS, Compose previews, and Android lifecycle integration (Activity, Fragment interop). USE WHEN: user mentions "Jetpack Compose", "@Composable" in Android-only context, "ViewModel", "Hilt", "Navigation Compose", "Material You", "Dynamic Color", "Compose preview", "AndroidView", "rememberLauncherForActivityResult", "Compose Wear OS" DO NOT USE FOR: Cross-platform Compose - use `frontend-frameworks/compose-multiplatform` DO NOT USE FOR: Kotlin language fundamentals - use `languages/kotlin` DO NOT USE FOR: Android non-UI APIs (Keystore, NFC, etc) - use `mobile/android-native`

LLM-as-reranker patterns. RankGPT listwise / pairwise / pointwise prompting, sliding window for long candidate lists, cost-latency vs Cohere Rerank, calibration tricks, structured output for rank lists. Code uses the Anthropic SDK. USE WHEN: user mentions 'LLM reranker', 'RankGPT', 'listwise reranking', 'pairwise reranking', 'pointwise reranking', 'GPT reranker', 'Claude reranker' DO NOT USE FOR: standard cross-encoder reranking - use `rag/reranking`; fine-tuning cross-encoders - use `retrieval/cross-encoder-training`; ColBERT - use `retrieval/colbert-retrieval`

Expert knowledge in quality assurance gates, code quality standards, and automated checks. Use when enforcing quality standards.

Refactor PHP code to improve structure, readability, and maintainability. Use when asked to refactor, restructure, improve code design, or apply design patterns.

Comprehensive integration skill for building sites with SvelteKit 2, Svelte 5, and Tailwind CSS v4

OpenAI's model connecting vision and language. Enables zero-shot image classification, image-text matching, and cross-modal retrieval. Trained on 400M image-text pairs. Use for image search, content moderation, or vision-language tasks without fine-tuning. Best for general-purpose image understanding.

Use this skill when the user asks to "review paper quality", "check paper completeness", "validate paper structure", "self-review before submission", or otherwise mentions systematic paper-quality checking. It provides a comprehensive quality-assurance checklist for academic papers.

Use when receiving code review feedback, before implementing suggestions—especially if the feedback seems unclear or technically questionable. This requires technical rigor and verification, not performative agreement or blind implementation.