メニュー
Use when generating images with Aliyun Bailian or DashScope Qwen-Image APIs, especially when you need region-aware HTTP calls, async task polling, or downloading temporary image URLs.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Search the web using Baidu AI Search Engine. Use for live information, documentation, or research topics.
Automatically optimize and split large files that exceed 600 lines. Use when a file is too large (over 600 lines), when you need to decouple a monolithic module, when a .tsx/.ts or .py file has mixed concerns (types, utils, hooks/components, CLI handlers, formatters), or when splitting a screen component or service module into smaller files. Applies to React/TypeScript frontends and Python backend modules.
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.
Use when you need a placeholder description. Replace with what the skill does and when an agent should trigger it.
Create and manage Python backend services using uv, FastAPI, Pydantic, SQLAlchemy, and AI libraries. Use this skill when the user asks to build a backend, API, or server-side application.
Create and manage modern React web applications using Bun, Vite, React 19, Tailwind CSS 4, and shadcn/ui. Use this skill when the user asks to build a frontend, a web app, or modify UI components.
| name | aliyun-image-gen |
| description | Use when generating images with Aliyun Bailian or DashScope Qwen-Image APIs, especially when you need region-aware HTTP calls, async task polling, or downloading temporary image URLs. |
使用阿里云百炼 Qwen-Image 文生图接口生成图片，默认走官方 HTTP API，不依赖 DashScope SDK。
先配置环境变量：
export ALIYUN_API_KEY="sk-xxx"
兼容旧变量名 DASHSCOPE_API_KEY，但优先读取 ALIYUN_API_KEY。
同步生成一张图并下载：
skills/aliyun-image-gen/scripts/aliyun_image_gen.py generate \
"一张极简风格的科技海报，白底，中心是蓝色折纸鹤" \
--model qwen-image-2.0-pro \
--size 1024*1024 \
--download
异步提交并等待完成：
skills/aliyun-image-gen/scripts/aliyun_image_gen.py generate \
"一张写实风格的咖啡馆照片，暖色夕阳光线" \
--mode async \
--model qwen-image-plus \
--wait \
--download
按任务 ID 查询异步结果：
skills/aliyun-image-gen/scripts/aliyun_image_gen.py status <task_id> --download
beijing；新加坡需显式传 --region singapore。qwen-image-2.0-pro，适合大多数文生图调用。qwen-image-plus 和 qwen-image，脚本会先做本地校验。--download。--negative-prompt: 反向提示词。--size: 输出分辨率，格式如 1024*1024、1664*928。--image-count: 期望输出张数；实际是否支持以模型约束为准。--no-prompt-extend: 关闭模型自动扩写提示词。--watermark: 打开水印，默认关闭。--json: 输出完整 JSON，方便二次处理。