imagen-qwen
agent
Uyari
Health Uyari
- License — License: MIT
- Description — Repository has a description
- Active repo — Last push 0 days ago
- Low visibility — Only 8 GitHub stars
Code Gecti
- Code scan — Scanned 1 files during light audit, no dangerous patterns found
Permissions Gecti
- Permissions — No dangerous permissions requested
Purpose
This is a Python-based agent skill designed for Claude Code and similar AI CLIs that enables natural language image generation. It uses the Alibaba Cloud Bailian API (Qwen/Wan models) to create images, with a specific focus on Chinese text rendering and photorealistic outputs.
Security Assessment
The overall risk is rated as Low. The automated code scan checked the core files and found no dangerous patterns, hardcoded secrets, or malicious code. The tool requires a third-party API key (`DASHSCOPE_API_KEY`) that you must provide and store as an environment variable. It makes external network requests to Alibaba Cloud to generate the images, which is expected behavior for this type of service. It does not request dangerous system permissions and does not appear to execute arbitrary shell commands beyond its intended Python script usage.
Quality Assessment
The project is very new and currently has low community visibility with only 8 GitHub stars, meaning it has not been widely battle-tested by a large user base. However, the repository is actively maintained (last updated today), has a clear description, and is fully licensed under the permissive and standard MIT license. The documentation is highly detailed, providing transparent instructions for installation, usage, and multi-region setup.
Verdict
Safe to use, though developers should be aware of the project's low community adoption and ensure they follow data handling best practices when configuring their Alibaba Cloud API keys.
This is a Python-based agent skill designed for Claude Code and similar AI CLIs that enables natural language image generation. It uses the Alibaba Cloud Bailian API (Qwen/Wan models) to create images, with a specific focus on Chinese text rendering and photorealistic outputs.
Security Assessment
The overall risk is rated as Low. The automated code scan checked the core files and found no dangerous patterns, hardcoded secrets, or malicious code. The tool requires a third-party API key (`DASHSCOPE_API_KEY`) that you must provide and store as an environment variable. It makes external network requests to Alibaba Cloud to generate the images, which is expected behavior for this type of service. It does not request dangerous system permissions and does not appear to execute arbitrary shell commands beyond its intended Python script usage.
Quality Assessment
The project is very new and currently has low community visibility with only 8 GitHub stars, meaning it has not been widely battle-tested by a large user base. However, the repository is actively maintained (last updated today), has a clear description, and is fully licensed under the permissive and standard MIT license. The documentation is highly detailed, providing transparent instructions for installation, usage, and multi-region setup.
Verdict
Safe to use, though developers should be aware of the project's low community adoption and ensure they follow data handling best practices when configuring their Alibaba Cloud API keys.
Claude Code skill for AI image generation using Alibaba Cloud Bailian API (通义千问/通义万相)
README.md
Imagen-Qwen — 阿里云百炼 AI 图像生成技能
自然语言生成高质量图像的 Claude Code / OpenClaw 技能,支持中文文字渲染和写实摄影。
为什么选择这个技能?
| 特性 | 本技能 | 原生 Claude Code | 其他图像技能 |
|---|---|---|---|
| 中文文字渲染 | ✓ 通义千问专项优化 | ✗ 无图像生成能力 | 部分支持 |
| 写实摄影级图像 | ✓ 通义万相多模型 | ✗ 无图像生成能力 | 部分支持 |
| 多模型可选 | ✓ 8 个模型按需切换 | ✗ 不适用 | 通常单模型 |
| 多尺寸预设 | ✓ 7+ 尺寸比例 | ✗ 不适用 | 部分支持 |
| 负面提示词 | ✓ 精细控制 | ✗ 不适用 | 部分支持 |
| 命令行直接调用 | ✓ 脚本即用 | ✗ 不适用 | 需自行编写 |
| 多区域 API | ✓ 中国/新加坡/弗吉尼亚 | ✗ 不适用 | 通常单区域 |
核心优势:
- 中文文字最佳 — 通义千问是目前在图像上渲染中文效果最好的模型之一
- 写实+艺术兼备 — 通义万相系列覆盖从快速草稿到专业级输出
- 即装即用 —
pip install两个包 + 一个 API 密钥即可开始
特性
- 通义千问 (Qwen-Image): 擅长在图像上渲染复杂的中英文文本
- 通义万相 (Wan Series): 写实图像和摄影级视觉效果
- 多种尺寸预设: 1:1, 16:9, 9:16, 4:3, 3:4 等
- 跨平台: 支持 Windows, macOS, Linux
- 多区域 API: 中国(默认)、新加坡、弗吉尼亚
安装技能
Claude Code(全局):
git clone https://github.com/Agents365-ai/imagen-qwen.git ~/.claude/skills/imagen-qwen
Claude Code(仅当前项目):
git clone https://github.com/Agents365-ai/imagen-qwen.git .claude/skills/imagen-qwen
OpenClaw:
git clone https://github.com/Agents365-ai/imagen-qwen.git skills/imagen-qwen
SkillsMP: 在 skillsmp.com 搜索 imagen-qwen,一键安装。
系统要求
环境要求
- Python 3.8+
- pip
安装依赖
pip install dashscope requests
API 密钥
从阿里云百炼控制台获取 API 密钥
export DASHSCOPE_API_KEY="your_api_key"
可选环境变量
# 设置默认模型(默认: qwen-image-plus)
export DASHSCOPE_MODEL="wan2.6-t2i"
# 设置 API 端点(默认: cn)
export DASHSCOPE_API_BASE="cn" # 或 "sg", "us", 或完整 URL
快速开始
自然语言(Claude Code)
直接告诉 Claude 你想要什么:
生成一只可爱的橘猫图片
创建一张带有"新年快乐"文字的海报
用 wan2.6-t2i 模型生成一张写实的山间日落照片
生成一张 16:9 的风景壁纸
命令行
# 基本用法(默认模型: qwen-image-plus)
python ~/.claude/skills/imagen-qwen/scripts/generate_image.py "一只可爱的猫咪" output.png
# 使用 Wan 模型生成写实图像
python ~/.claude/skills/imagen-qwen/scripts/generate_image.py --model wan2.6-t2i "山间日落" photo.png
# 自定义尺寸
python ~/.claude/skills/imagen-qwen/scripts/generate_image.py --size 16:9 "宽屏风景" landscape.png
# 使用负面提示词
python ~/.claude/skills/imagen-qwen/scripts/generate_image.py --negative "模糊" "高质量人像" portrait.png
# 列出可用模型
python ~/.claude/skills/imagen-qwen/scripts/generate_image.py --list-models
模型
| 模型 | 最佳用途 |
|---|---|
qwen-image-plus |
中英文文本、海报、插画 |
wan2.6-t2i |
写实照片、人像 |
wan2.5-t2i-preview |
高质量艺术作品 |
wan2.2-t2i-flash |
快速生成 |
wan2.2-t2i-plus |
专业级 |
wanx2.1-t2i-turbo |
快速执行 |
wanx2.1-t2i-plus |
专业级 |
wanx2.0-t2i-turbo |
早期版本 |
尺寸预设
通义千问:
1:1→ 1024×102416:9→ 1664×9289:16→ 928×16644:3→ 1216×9123:4→ 912×1216
通义万相:
1:1→ 1024×10241:1-large→ 1280×128016:9→ 1280×7209:16→ 720×12804:3→ 1200×9003:4→ 900×12002:1→ 1440×720
API 端点
| 地区 | 别名 | URL |
|---|---|---|
| 中国(默认) | cn |
https://dashscope.aliyuncs.com/api/v1 |
| 新加坡 | sg |
https://dashscope-intl.aliyuncs.com/api/v1 |
| 弗吉尼亚 | us |
https://dashscope-us.aliyuncs.com/api/v1 |
许可证
MIT License
支持
如果这个项目对你有帮助,欢迎支持作者:
微信支付 |
支付宝 |
Buy Me a Coffee |
作者
Agents365-ai
- Bilibili: https://space.bilibili.com/441831884
- GitHub: https://github.com/Agents365-ai
Yorumlar (0)
Yorum birakmak icin giris yap.
Yorum birakSonuc bulunamadi