← Back to Skills Marketplace
joannaxing

Doubao Maliang Image Gen

by JoannaXing · GitHub ↗ · v1.0.1 · MIT-0
cross-platform ✓ Security Clean
213
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install doubao-maliang-image-gen
Description
小马良-豆包生图。Generate images with Doubao Seedream via Volcano Engine ARK. Supports Seedream 5.0 and other models. Use when the user invokes 小马良 or requests Douba...
README (SKILL.md)

小马良-豆包生图 / Maliang Doubao Image Gen

🖌️ 关于「马良」: 名字源自中国神话故事《神笔马良》—— 马良拥有一支神笔,画什么就会变成真的。这个 skill 就是你的「神笔」,用文字描绘,让 AI 为你生成图像。

🖌️ About "Maliang": Named after the Chinese legend Shenbi Maliang (Magic Brush Ma Liang) — who possessed a magic brush that brought drawings to life. This skill is your magic brush: describe with words, let AI generate the image.

基于火山引擎方舟 ARK 平台的豆包 Seedream 文生图服务。

Powered by Doubao Seedream text-to-image service via Volcano Engine ARK platform.

默认模型 / Default Model: doubao-seedream-5-0-260128 (Seedream 5.0)


✨ 核心特色 / Key Features

  1. 一键生图,自动回传 / One-Click Generation, Auto-Reply

    在聊天窗口直接呼唤马良,描述你想要的图片,生成完成后自动将图片发送回聊天界面,无需手动查找文件。

    Simply invoke Maliang in chat, describe what you want, and the generated image is automatically sent back to the chat window — no need to manually locate files.

  2. 本地备份 + 画廊 / Local Backup + Gallery

    图片同时保存到本地,并生成可视化画廊页面,方便管理和回顾。

    Images are also saved locally with a visual gallery page for easy management and review.


🚀 快速开始 / Quick Start

1. 配置 API Key / Configure API Key

推荐:使用 Skill 环境变量 / Recommended: Skill Environment Variables

在 OpenClaw 配置中为该 skill 设置环境变量: Set environment variables in OpenClaw config for this skill:

{
  "skills": {
    "doubao-maliang-image-gen": {
      "env": {
        "VOLCANO_ENGINE_API_KEY": "your-api-key-here"
      }
    }
  }
}

文件位置 / File location: ~/.openclaw/config.json

替代方式:系统环境变量 / Alternative: System Environment Variables

export VOLCANO_ENGINE_API_KEY="your-api-key-here"

兼容的变量名 / Compatible Variable Names(按优先级 / in priority order):

  • VOLCANO_ENGINE_API_KEY (推荐 / recommended)
  • ARK_API_KEY
  • SEEDREAM_API_KEY

⚠️ 安全提示 / Security Note: 永远不要将 API key 直接写在命令行参数或脚本里。使用环境变量避免密钥泄露。

Never write API keys directly in command line arguments or scripts. Use environment variables to prevent credential leaks.

2. 获取 API Key / Get API Key

中文步骤:

  1. 访问 火山方舟控制台
  2. 创建或进入已有项目
  3. 在「API Key 管理」中创建新密钥
  4. 开通「图像生成」权限(Doubao Seedream 模型)

English Steps:

  1. Visit Volcano Engine ARK Console
  2. Create or enter an existing project
  3. Create a new key in "API Key Management"
  4. Enable "Image Generation" permission for Doubao Seedream model

🎯 触发条件 / When to Use

使用本 skill 的场景(中英文均可触发): Use this skill when (works in both Chinese and English):

  • 用户点名 小马良 / User invokes Maliang

    • 可触发关键词 / Trigger words: "小马良", "马良", "Maliang", "Little Maliang"
  • 用户明确要求使用 豆包 / Doubao / Seedream / 火山方舟 / Volcano ARK 生图

    • 可触发关键词 / Trigger words: "豆包", "Doubao", "Seedream", "火山方舟", "Volcano ARK", "Volcano Engine"
  • 用户希望图片生成走 Maliang 路径 / User requests the Maliang image generation path

💡 提示 / Tip: 无论用中文还是英文呼唤,马良都能听懂!

Maliang understands both Chinese and English invocations!


💬 使用示例 / Usage Examples

中文示例:

用户:小马良,帮我画一只穿着宇航服的猫咪,在月球上散步,卡通风格

马良:✨ 已为你生成图片!(图片自动发送到聊天窗口)

English Example:

User: Maliang, draw me a cat in an astronaut suit walking on the moon, cartoon style

Maliang: ✨ Here's your image! (Image automatically sent to chat)

更多示例 / More Examples:

场景 / Scenario 对话示例 / Chat Example
产品设计 / Product Design "马良,帮我生成一个极简风格的咖啡杯产品图,白色背景,柔和光线"
插画创作 / Illustration "小马良,画一幅赛博朋克风格的城市夜景,霓虹灯,雨天街道"
头像生成 / Avatar "马良,给我生成一张卡通风格的程序员头像,戴眼镜,友善的表情"
概念图 / Concept Art "帮我画一片秋日森林,金色阳光透过树叶,风景摄影风格"

💡 Prompt 建议 / Prompt Tips

豆包 Seedream 支持多种风格描述,建议包含: Doubao Seedream supports various style descriptions. Recommended elements:

要素 / Element 说明 / Description 示例 / Example
主体 / Subject 画面中有什么 / What's in the image a vintage sports car
场景 / Scene 环境、背景 / Environment, background on a coastal highway at dusk
风格 / Style 摄影、插画、3D 等 / Photography, illustration, 3D, etc. cinematic photography
构图 / Composition 近景、全景等 / Close-up, wide shot, etc. wide-angle shot
光线/色调 / Lighting/Color 光线、色彩 / Light, color scheme golden hour lighting, warm tones
质感 / Texture 材质 / Material quality glossy metallic finish

示例 / Example:

"Editorial fashion portrait, model in seafoam silk dress, gold jewelry, soft natural window light, clean beige background, high-end magazine style"


📤 输出说明 / Output Structure

图片会同时输出到两个地方: Images are delivered to two places simultaneously:

  1. 聊天窗口 / Chat Window — 生成完成后自动发送到当前对话 ✨

  2. 本地文件夹 / Local Folder

~/.openclaw/workspace/tmp/doubao-maliang-image-gen-\x3Ctimestamp>/
├── \x3Cprompt-slug>-1.png    # 生成的图片 / Generated image
├── manifest.json          # 生成记录(包含 prompt、模型、时间戳)/ Generation record
└── index.html             # 本地画廊页面 / Local gallery page

🔧 技术细节 / Technical Details

命令行生成 / CLI Generation

如果需要通过命令行直接生成(而非聊天触发): If you need to generate via command line (instead of chat invocation):

基础用法 / Basic Usage:

python3 "$SKILL_DIR/scripts/gen.py" --prompt "a serene mountain landscape at sunset"

多图生成 / Multiple Images:

python3 "$SKILL_DIR/scripts/gen.py" --prompt "minimalist coffee cup product shot" --count 2

指定尺寸 / Specify Size:

python3 "$SKILL_DIR/scripts/gen.py" --prompt "cyberpunk city street" --size 1K
# 支持 / Supported: 1K (1024x1024), 2K (2048x2048)

切换模型 / Switch Models

通过环境变量(全局)/ Via Environment Variables (Global)

export SEEDREAM_MODEL="doubao-seedream-5-0-260128"
export SEEDREAM_API_ENDPOINT="https://ark.cn-beijing.volces.com/api/v3/images/generations"

通过命令行参数(单次)/ Via Command Line (One-time)

python3 "$SKILL_DIR/scripts/gen.py" \
  --prompt "abstract art composition" \
  --model "doubao-seedream-5-0-260128" \
  --endpoint "https://ark.cn-beijing.volces.com/api/v3/images/generations"

支持的模型 / Supported Models

模型 ID / Model ID 说明 / Description
doubao-seedream-5-0-260128 Seedream 5.0 (默认 / default)

如需使用其他模型,请先在方舟控制台开通对应模型的调用权限。

To use other models, please enable the corresponding model permissions in the ARK console first.

API 信息 / API Info

  • API 端点 / API Endpoint: https://ark.cn-beijing.volces.com/api/v3/images/generations
  • 协议 / Protocol: OpenAI-compatible HTTP API
  • 默认尺寸 / Default Size: 2K (2048x2048)
  • 单次最大数量 / Max per Request: 4 张 / images
  • 超时设置 / Timeout: 300 秒 / seconds

🐛 故障排除 / Troubleshooting

问题 / Issue 解决方案 / Solution
"Missing API key" 检查 VOLCANO_ENGINE_API_KEY 环境变量是否设置 / Check if VOLCANO_ENGINE_API_KEY env var is set
"API failed (401)" API Key 无效或过期,检查密钥是否正确 / API Key invalid or expired
"API failed (429)" 请求过于频繁,稍后重试 / Too many requests, retry later
"no images returned" 检查 prompt 是否为空或包含违规内容 / Check if prompt is empty or contains prohibited content

License

MIT

Usage Guidance
This skill appears to be what it claims: a small Python client that calls Volcano Engine Seedream, saves images locally, and prints paths. Before installing: 1) Only provide a Volcano Engine API key you trust and scope that key to image-generation permissions; 2) avoid pointing SEEDREAM_API_ENDPOINT to unknown servers (overriding the endpoint could leak your API key/payloads); 3) note that the script saves files under ~/.openclaw/workspace/tmp and the gallery index.html may contain your prompts (the code escapes HTML); 4) if you need stricter containment, run the skill in an isolated environment or use a minimally privileged API key. If you want absolute assurance, review the script locally (scripts/gen.py) and test with a limited/sandbox API key first.
Capability Analysis
Type: OpenClaw Skill Name: doubao-maliang-image-gen Version: 1.0.1 The skill is a well-implemented tool for generating images via the Volcano Engine ARK API (Doubao Seedream). The Python script (scripts/gen.py) uses standard libraries, implements proper input sanitization for filenames (slugify), and escapes HTML content for the local gallery output to prevent XSS. No evidence of data exfiltration, malicious execution, or prompt injection was found.
Capability Assessment
Purpose & Capability
Name/description (Doubao Seedream via Volcano Engine) align with the requested binary (python3), the single required env var (VOLCANO_ENGINE_API_KEY), and the included Python client that POSTs to a Volcano ARK images endpoint.
Instruction Scope
SKILL.md and the script focus on prompting the Seedream API and saving images locally. The doc promises 'automatic send back to the chat window' — the included script only prints file paths and saves files; sending into the chat is likely performed by the OpenClaw platform integration rather than the script. Otherwise the runtime instructions do not request unrelated files, credentials, or system state.
Install Mechanism
No install spec; this is an instruction-only skill with a small Python script. Nothing is downloaded or written outside the user's workspace/home by an installer.
Credentials
Requires a single API key (VOLCANO_ENGINE_API_KEY) which is appropriate. The script also allows overriding the API endpoint via SEEDREAM_API_ENDPOINT (and accepts alternate key names); allowing an arbitrary endpoint is convenient but means a misconfigured endpoint could capture the API key or payloads — this is a user-configurable option, not required by default.
Persistence & Privilege
always is false and the skill does not request elevated system privileges. It writes outputs under ~/.openclaw/workspace/tmp which is proportional to its purpose.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install doubao-maliang-image-gen
  3. After installation, invoke the skill by name or use /doubao-maliang-image-gen
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.1
Add GitHub homepage URL to metadata
v1.0.0
Initial release: Doubao Seedream image generation with auto-reply to chat, local gallery, and bilingual support.
Metadata
Slug doubao-maliang-image-gen
Version 1.0.1
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 2
Frequently Asked Questions

What is Doubao Maliang Image Gen?

小马良-豆包生图。Generate images with Doubao Seedream via Volcano Engine ARK. Supports Seedream 5.0 and other models. Use when the user invokes 小马良 or requests Douba... It is an AI Agent Skill for Claude Code / OpenClaw, with 213 downloads so far.

How do I install Doubao Maliang Image Gen?

Run "/install doubao-maliang-image-gen" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Doubao Maliang Image Gen free?

Yes, Doubao Maliang Image Gen is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Doubao Maliang Image Gen support?

Doubao Maliang Image Gen is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Doubao Maliang Image Gen?

It is built and maintained by JoannaXing (@joannaxing); the current version is v1.0.1.

💬 Comments