360 Secure SkillHub
360 安全技能中心

markdown-to-minimal-image

核心功能:读取Markdown文档生成精准摘要,再生成极简手绘白板风格图片,支持指定比例(含5:2/5:4等社交横幅格式)及配色。典型场景:需把教程、笔记类Markdown转为适合社交平台、官网横幅的简约可视化内容。输入为Markdown文件,可选比例、配色;输出含文档摘要、生成的图片及保存路径等信息。

j3ffyang
j3ffyang
热度 14.50
最近更新 2026-08-08 14:13:54
分类:创作依赖:读写文件 / 内部调用大模型应用场景:图像生成 / 摘要总结 / 社交媒体

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add 260611-md2mini-img
Metadata
name
markdown-to-minimal-image
description
Read an attached Markdown document, analyze and summarize it, then generate a minimalist image from the summary using the default image model. Ask the user for aspect ratio when needed, and support social-banner formats such as 5:2 and 5:4. Use when the user wants a document turned into a simple tutorial-style visual with handwritten whiteboard aesthetics.
version
1.0.0
author
Jeff Yang (https://github.com/j3ffyang)
license
MIT
platforms
["linux","macos","windows"]
metadata
{"hermes":{"tags":["image-generation","markdown","summary","whiteboard","social-banner"],"requires_toolsets":["file"],"requires_tools":[],"required_environment_variables":[]}}

Markdown to Minimal Image

When to Use

Use this skill when the user provides a Markdown document and wants:

  • A concise summary of the document.
  • One generated image based on that summary.
  • A specific aspect ratio for social media or hero banners.
  • A minimalist, handwritten, whiteboard-like visual style.

When NOT to Use

Do NOT use this skill for:

  • Generating photorealistic images or detailed illustrations.
  • Creating multi-image documents or slide decks.
  • Processing non-Markdown file formats (PDF, Word, plain text without conversion).
  • Tasks where the user explicitly requests a different visual style (photography, 3D rendering, oil painting).
  • Generating images without a source document to summarize.

Inputs

  • document — Attached Markdown file to read and analyze.
    • Required: yes
    • Type: file (.md)
  • aspect_ratio — Desired image aspect ratio, e.g. 5:2, 5:4, 16:9, 1:1.
    • Required: no
    • Type: string
    • Default: none (user will be prompted)
  • color_scheme — Color scheme for the image, e.g. black-and-white, blue accent, warm beige, pastel.
    • Required: no
    • Type: string
    • Default: neutral minimalist palette

Outputs

  • summary — String. 3–5 bullet points or one short paragraph summarizing the document.
  • image — Generated image object. Exactly one image in the requested aspect ratio and color scheme.
  • image_path — String. File path where the image is saved: $HOME/.hermes/cache/{date}-{subject}.png.
  • aspect_ratio — String. The aspect ratio used for generation (resolved from input or user prompt).
  • color_scheme — String. The color scheme used for generation (resolved from input or default).

Procedure

  1. Read the attached Markdown document using file tools.
  2. Detect the document language. If the document is in Simplified Chinese, all output (summary, image text) must be in Simplified Chinese. If English, all output must be in English.
  3. Identify the core topic, key ideas, and actionable points.
  4. Summarize the document in 3–5 short bullets or one short paragraph.
  5. If aspect_ratio is missing, ask the user for it before proceeding. Accept only valid ratios (5:2, 5:4, 16:9, 1:1, or a custom ratio the user specifies).
  6. If color_scheme is missing, ask the user or default to a neutral minimalist palette.
  7. Build an image prompt from the summary that incorporates the aspect ratio, color scheme, and whiteboard-tutorial style.
  8. Generate exactly one image using the default image generation model.
  9. Save the generated image to $HOME/.hermes/cache/{date}-{subject}.png.
  10. Call present_files to surface the saved image to the user.
  11. Return the summary alongside the image.

Verification

  • The summary must be non-empty and accurately reflect the source document.
  • The generated image must match the requested aspect ratio exactly.
  • The generated image must use the requested color scheme (or default).
  • The image style must be minimalist, whiteboard-like, with handwritten text — not photorealistic or cluttered.
  • The language of the image text must match the source document language (English or Simplified Chinese).

Error Handling

  • Missing document: If no Markdown file is attached, inform the user and halt.
  • Invalid aspect ratio: If the user provides an unsupported ratio, list valid options and ask again.
  • Image generation failure: If the model fails to produce an image, retry once with a simplified prompt. If still failing, report the error and suggest a different model or lower resolution.
  • Unreadable document: If the file cannot be parsed as Markdown, inform the user and request a valid .md file.

Style Notes

  • Handwritten tutorial on whiteboard.
  • Minimalist composition with clean layout.
  • Simple color palette matching the requested scheme.
  • Low resolution acceptable for social media or hero-banner use.
  • Prefer simple diagrams, arrows, boxes, labels, and short notes.
  • Avoid heavy detail, photorealism, or decorative clutter.

Supported Aspect Ratios

  • 5:2 — social banner
  • 5:4 — social banner
  • 16:9 — hero / widescreen
  • 1:1 — square / social post
  • Custom ratios accepted on request