Explore skills

Use when the user needs to write, polish, or verify Chinese academic reports (思政报告, 国家安全报告, 博雅报告, 活动报告, 读书报告, 社会实践报告). This skill also handles requests for checking AI-generated content, reducing AI writing traces, or automatically generating reports from reference materials, routing to specific AI-related skills based on intent.

Formats approved test artifacts into review-ready docs and importable outputs for Gherkin, Xray, Zephyr, TestLink, and TestRail.

Use when starting a new code project that needs structured task management, RIPER-7 lifecycle, version planning, session handoff, and daily reporting. Triggers include "init workflow", "bootstrap .tasks", "initialize project workflow", "新项目初始化工作流", "初始化任务工作流".

Generates DIN 5008-compliant business letters as PDFs. Optionally, sends them via the LetterXpress Print-and-Mail API (test mode).

Query local LM Studio or any OpenAI-compatible LLM server (like Ollama, llama.cpp, vLLM) for coding tasks, explanations, or text generation without using paid API tokens. Use when the user explicitly asks to use a local model, or for low-stakes tasks like code examples, documentation, simple scripts, or exploratory work where perfect accuracy isn't critical.

Use this tool when creating, reviewing, or iterating on any user interface, including React components, HTML pages, mobile screens, wireframes, or Figma screenshots. It's also valuable for self-evaluating UI code and addressing questions about usability or UX quality, by running a five-layer Body Test that simulates human interaction.

Initializes a Vue3 project (Router, Pinia, ESLint, Prettier, Element Plus) supporting JS/TS with an optional package manager. Use --ts for TypeScript and --no-axios to skip Axios encapsulation.

Generate images, videos, and audio using the fal.ai API. Use when the user works with fal.ai, fal_client, @fal-ai/client, Flux Pro, Kling video, or mentions AI image/video generation with fal.

Injects the current system date, time, day of week, and timezone into the conversation. This skill is used for time-related questions and proactively at the start of new conversations to establish time context.

Control a local ComfyUI instance via its REST API to generate AI images, videos, and audio. Use this skill for ComfyUI mentions, AI media generation, Stable Diffusion workflows, and managing generation tasks or models.

Keep CLAUDE.md documentation in sync with codebase evolution. Use this skill when the project structure, tech stack, or architecture has changed and documentation may be outdated, or when the user explicitly requests to sync, update, or check CLAUDE.md.

This skill creates discussion agendas, guides, or structured conversation guides for media episodes (podcasts, TV shows, or movies). It triggers on phrases like 'discussion agenda', 'discussion guide', or requests for structured conversation questions.