Skills — le catalogue de compétences de vos agents
Chaque skill est une capacité prête à l'emploi (GitHub, PDF, analyse d'images, marketing…) que votre agent charge et utilise de façon autonome. Installez en un clic depuis votre portail.
Upload image files to PixelVault and get a public CDN URL back. Use when the user says "screenshot this and host it", "upload this image", "host this mockup"...
List recently uploaded images on PixelVault with their CDN URLs. Use when the user asks to "see my uploads", "find that image URL", "what images have I hoste...
短视频爆款钩子生成器。输入视频主题或脚本，自动生成10种高点击率的开场钩子，涵盖"冲突型、数字型、悬念型、情绪型、揭秘型"等类型。当用户说"生成钩子"、"写视频开头"、"爆款标题"、"开头怎么写"、"生成吸引人的开头"时触发此技能。
Create reusable paged HTML tutorial guides from operation workflows, screenshots, videos, links, reference images, and step-by-step documentation. Use when t...
External-link index of major companies' design systems (Material, Fluent, Carbon, Polaris, Atlassian, Lightning), design-token tools and specs, pattern-libra...
Generate green-screen raster subject assets with an image model and convert them into transparent PNGs with robust chroma-key cutout. Use when Codex needs pe...
Index of ready-made design assets — stock photos, icon sets, fonts, color palettes, mockups, UI kits, and templates — sourced from awesome-design and design-...
Image generation and editing based on Google Gemini native image generation (Nano Banana).
当用户给出一个产品或界面设计想法时，自动按标准模版范式将其优化为结构清晰、细节完备的设计描述。适用于小程序、App、网页、H5 等任何界面类产品设计任务。
Generate beautiful Apple HIG-inspired HTML visualizations (bar, line, donut, progress, stat-card) and screenshot them as PNG images. Use when user asks to vi...
A design-language assistant for Ant Design (antd). Use when the user wants to learn, apply, critique, or theme UIs according to Ant Design's design language — its four design values
Use when building, implementing, or refactoring frontend UI — components, pages, layouts, interactions, or any visual work in pure HTML/CSS/JS. Use when inte...
Java 设计模式权威参考。涵盖 GoF 23 种经典模式及企业级架构模式的完整 Java 实现，含 26 个核心设计原则。当用户询问设计模式、架构模式、设计原则（KISS/DRY/SOLID/GRASP 等）、Java 模式实现、或需要选择合适的设计模式时调用此技能。覆盖创建型、结构型、行为型、架构型、并发型等 15 大类 182 个模式 + 26 个设计原则。
Plan subtitle specs before transcribing or burning captions — CPS limits, line breaks, timing gaps, SRT/VTT rules, SDH cues, and platform targets for readabl...
YouTube Channel Management: Upload File Manager videos, update video metadata, privacy, localizations, thumbnails, captions, playlists, playlist items, chann...
Public Commons Media Search: Search Wikipedia and Wikimedia Commons for free images, audio, video. Get file URLs and licensing info. Supports pagination. Use...
Real Estate Aerial Video Generator: Create cinematic aerial videos from street addresses, check whether a video is ready yet, and return downloadable MP4 lin...
本地图片视觉理解桥接。当用户明确要求分析、描述或理解图片内容时，调用本地 Ollama 视觉模型（qwen3.5:4b）读取图片并以文字描述返回，让不具备多模态能力的推理模型也能"看见"图片。
科幻小说写作助手。帮助用户从零开始构思、搭建世界观、设计角色、规划情节、撰写章节，支持完整长篇创作和短篇创作。当用户提到写科幻小说、科幻故事、赛博朋克、太空歌剧、硬科幻、软科幻、未来世界、外星文明、时间旅行、末日题材、星际战争、人工智能觉醒、虚拟现实等科幻创作相关需求时触发。也适用于用户说"帮我写个科幻短篇""构...
百度图片下载：按关键词从百度图片搜索批量下载图片到本地。触发场景：用户说『下载 关键词 X 张图片』『百度图片下载 X』『用百度抓 N 张图』『按关键词 抓百度图』『百度图片批量下载』『Baidu image download』。底层用 acjson JSON 端点（image.baidu.com/search/...
音乐库整理工作流：扫描→规划→主人确认→软删除→验证→清理。强调安全流程。Safe-by-default music library organizer with soft-delete + mandatory owner confirmation.
Elite frontend image-direction skill for generating premium, conversion-aware website design references. CRITICAL OUTPUT RULE — generate ONE separate horizon...
Elite mobile app image-generation skill for creating premium, app-native screen concepts and flows. Designed for iOS, Android, and cross-platform mobile prod...
Practical guidance for training MoE VLMs in Megatron Bridge. Compares FSDP and 3D-parallel approaches, using rounded lessons from Qwen3-VL, Qwen3-Next, and other multimodal experiments.