Explorar skills

This skill should be used when the user asks to "write a Seedance prompt", "create a video with Seedance", "generate Seedance 2.0 prompt", "use Seedance for video generation", "extend a video", "edit a video with Seedance", "do music sync video", "replicate camera movement", "make an AI video", or mentions Seedance, 即梦, video generation prompts, multi-modal video creation, or AI video editing. Pro

Ship a complete production-ready feature end-to-end while the user sleeps or is away. Reads vision/context from the user, runs /autoplan, fans out aggressive parallel research subagents (context7, Nia, general-purpose) only when needed, executes test-driven implementation with /frontend-design for any UI, runs adversarial break-it testing, full prod-readiness checklist (security, perf, observabili

Maintain session work continuity across Claude.ai rate-limit boundaries. Auto-pauses when utilization reaches a threshold, sleeps in 1-hour chunks via ScheduleWakeup, then resumes the same task in the same conversation. MUST be invoked via /loop dynamic mode (e.g. `/loop /continuous 95 "<task>"`); refuses otherwise.

智能 Token 优化工具，通过分析、压缩、缓存、预算执行四大策略显著降低 AI 代理的 Token 消耗。 核心差异化：业界唯一可在执行前检查预算并拦截超预算调用的方案。 使用场景：token成本控制、预算管理、成本监控、agent优化。

Generate professional AI images with Hypereal's Nano Banana 2 model. Add your image ideas to concepts.txt -- Gemini Flash writes expert prompts, Hypereal generates the images. Run from the folder containing generate_prompts.py, generate_images.py, and .env.

Extract personal data (notes, transcripts, AI summaries, attachments, custom recipes, chat history) from Granola.ai via their undocumented REST API using locally-stored auth tokens. Use when the user wants to back up, recover, snapshot, or export their Granola data — especially after a tier downgrade where the UI hides data the API still serves. Triggers on "export granola", "backup granola", "ext

Use this skill whenever the user wants to brainstorm or plan a project, app, feature, or tool — even casually ("I want to build X", "help me think through Y", "how should I approach Z"). Generates a structured 3-stage prompt file plan: architecture+MVP, core feature polish, and final touch-up. Trigger on any project ideation or "where do I start" type request.

Cognitive Debt Guard - Prevent the 23.5% incident spike from AI-generated code. Comprehension gates, review frameworks, and AI-free zones. Based on 2026 research.

Maximum reasoning + multi-agent orchestration for any task. Wraps your request with ultrathink deep reasoning, parallel agent execution, and a fresh-eyes quality gate. Adaptive: scales agent count to task complexity. Works for code, writing, analysis, debugging, and research. Trigger: "deep work", "deep mode", "max mode"

Create S-tier portable skills with comprehensive quality guidance. This skill should be used when creating new skills, improving existing skills, auditing skill quality, or learning skill-building best practices. Covers the full lifecycle from use case definition through testing, distribution, and iteration. Do NOT use for general coding tasks, writing documentation unrelated to skills, or buildin

Ruthless automated code review for any tech stack. 31+ frameworks, post-mortem verification, game dev support, anti-AI-slop detection.

유튜브 레퍼런스 영상 분석 → 벤치마킹 → 30%+ 각색 → 스토리보드 기획 → rfn-video-pipeline용 JSON 생성 → 이미지 생성 → HTML 시각화 → AI 영상 제작까지의 전체 쇼츠 제작 워크플로우. 유튜브 쇼츠 분석, 벤치마킹, 각색, 스토리보드, 레퍼런스 영상, 바이럴 분석, 감정 공학, 쇼츠 팩토리, 영상 기획, 콘텐츠 클론 시 자동 활성화.