auto-article-images 智能配图助手

SkillMedia

Your AI can go through an article or draft, work out which parts need images and what each one should show, and turn that into an actionable result. auto-article-images is a skill for content creators, operations teams, technical writers, and media publishers who find adding images to articles a hassle. It works from the material you provide, producing summaries, recommendations, and reusable deliverables along the way.

Available today. Use it from your connected AI after setup.

Add the skill, then share the article or draft you want illustrated and ask your AI to work out the images it needs. Everything is built from the material you provide, so include the text you are working from.

Then ask your AI: use the auto-article-images 智能配图助手 skill

What your AI can do with it

  • Identify where an article needs images
  • Suggest what each image should show based on your content
  • Break the image work into clear steps and flag which images matter most
  • Summarize your material and diagnose what is missing
  • Recommend concrete next steps for getting the images done
  • Turn your material into reusable deliverables

What this skill tells your AI

The instructions your AI receives, as published by allinherog-star/ai-skills in skills/auto-article-images/SKILL.md and read by ahel’s review.

快速开始

更多技能

概述

智能配图助手用于回答「还要配图,好麻烦」、配图、封面、文章图片,适合内容创作者、运营、technical、内容媒体在明确业务目标、内容材料或分析对象后调用。 它会结合正文、文章正文或 Markdown等输入,整理关键上下文,并输出摘要、诊断结论、行动建议和可复用交付物,便于继续执行、复盘或交付。

什么时候使用

适用场景

  • 用户提出“还要配图,好麻烦”这类问题,需要快速拆解目标、判断重点并形成可执行结果
  • 内容创作者、运营、technical、内容媒体需要围绕智能配图助手生成摘要、诊断结论、行动建议和可复用交付物
  • 用户已经准备了视觉风格、标题、配图数量,希望整理成可执行的分析或优化结果
  • 用户需要把智能配图助手相关材料转成清晰结论、优先级和下一步动作

调用方式

通过导出的 Python runner 直接调用 AI Skills API:

命令示例

基础调用

python3 scripts/run.py --params '{}'

带常用参数调用

python3 scripts/run.py --params '{"style":"clean-commercial"}'

参数说明

参数类型必填默认说明
stylestringclean-commercial视觉风格;可选值:清爽商业(clean-commercial)、杂志摄影(editorial-photo)、柔和插画(soft-illustration)、科技图解(tech-diagram
titlestring-标题
imageCountinteger4配图数量
sourceTextstring-文章正文或 Markdown
coverRatiosarray["2.35:1","1:1","16:9","3:4","9:16"]封面比例
watermarkModestringoff水印;可选值:不加(off)、右下角(corner)、平铺(tiled)、两者(both
sourceDocumentstring-支持 docx、md、txt 等文本文件
publishPlatformstringall发布平台;可选值:全平台(all)、公众号(wechat)、小红书(xhs)、博客(blog)、知乎(zhihu
brandRequirementsstring-品牌要求

完整机器可读参数结构见 references/form-schema.json

参数取值参考

当前技能没有需要额外查表的分类参数。

支持的输入格式

当前技能直接接收 JSON 参数,不涉及分享链接解析。

示例请求

下面的示例参数可直接传给 scripts/run.py,runner 会把它们发送给 AI Skills API。

python3 scripts/run.py --params '{"style":"clean-commercial"}'

等价的 --params JSON:

{
  "style": "clean-commercial"
}

返回结果示例

{
  "success": true,
  "data": {
    "mode": "async",
    "status": "completed",
    "resultEnvelope": {
      "status": "completed",
      "title": "配图结果已生成",
      "summary": "已生成正文配图和多比例封面。",
      "items": [
        {
          "id": "image-package",
          "type": "attachmentGroup",
          "title": "配图文件",
          "artifactIds": [
            "content-01.webp",
            "cover-wechat-2_35x1.webp"
          ]
        }
      ],
      "artifacts": [
        {
          "id": "content-01.webp",
          "name": "content-01.webp",
          "relativePath": "content-01.webp",
          "mimeType": "image/webp",
          "url": "/api/skill-artifacts/job_demo/file/content-01.webp"
        },
        {
          "id": "cover-wechat-2_35x1.webp",
          "name": "cover-wechat-2_35x1.webp",
          "relativePath": "cover-wechat-2_35x1.webp",
          "mimeType": "image/webp",
          "url": "/api/skill-artifacts/job_demo/file/cover-wechat-2_35x1.webp"
        }
      ],
      "presentation": {
        "mode": "single"
      }
    },
    "zipUrl": "/api/skill-artifacts/job_demo/archive"
  },
  "meta": {
    "executionTime": 842,
    "cached": false
  }
}

交付内容

  • 正文配图:围绕文章主题生成可插入内容中的图片素材。
  • 多比例封面:适配不同内容平台的封面使用场景。
  • 图片交付包:方便继续下载、筛选、发布或二次设计。

结果使用建议

  • 先看图片是否准确表达文章主题,再选择适合正文插入和封面展示的素材。
  • 不同平台对封面比例和视觉风格要求不同,使用前建议按发布渠道做最终选择。
  • 输入文章摘要、目标平台和风格偏好越清晰,生成结果越容易贴合内容。

运行前准备

  • AISKILLS_BASE_URL:默认 https://ai-skills.ai
  • AISKILLS_API_KEY:必填,用于认证调用
  • AISKILLS_TENANT_ID:默认 default

Signals

GitHub stars
51
Forks
5
Last commit
Jun 2026
Advanced
Catalog kind
skill
Gateway key
auto-article-images
Source
github.com/allinherog-star/ai-skills