shaom/infocard-skills

Registry 信息

项目简介

infocard-skills 是一个开源的代理技能集合，能够将自然语言输入转化为编辑风格的信息卡片。专为 AI 代理设计，可自动生成视觉结构清晰、可直接发布的资讯卡片，无需手动设计。

要点

• 从自然语言生成编辑风格信息卡片
• 开源的 AI 代理技能，便于集成
• 输出结构清晰、可直接发布的卡片
• 轻量级，易于部署

使用场景

• AI 代理自动生成新闻摘要卡片
• 快速制作编辑内容预览
• 聊天机器人中将数据转换为视觉卡片
• 快速原型设计信息展示

README 摘要

[中文](./README.zh-CN.md) # infocard-skills Open-source agent skills for generating editorial-style information cards from natural-language input.  `infocard-skills` is a small repository of installable skill packages for agents. The current package, `editorial-card-screenshot`, turns source material into screenshot-ready HTML cards and optional PNG outputs. ## Install Install the full skill package, not a single file: - [`skills/editorial-card-screenshot`](./skills/editorial-card-screenshot) Copy or install the entire directory into your agent's skills location. Keep these files together: - `SKILL.md` - `assets/` - `references/` - `scripts/` - `agents/` when supported by the host ## Verify After installation, test with one of these prompts: - `Use $editorial-card-screenshot to turn these notes into an editorial info card.` - `Use $editorial-card-screenshot to make a 3:4 information card from this text.` - `Use $editorial-card-screenshot and give me both the HTML and PNG.` ## Usage Typical requests: - `Turn this article into an editorial information card.` - `Make a portrait 3:4 info card from these notes.` - `Summarize this content as a dense magazine-style card.` - `Create the HTML only. I do not need the image.` ## Render Requirement For PNG capture, the current helper expects local Google Chrome at: ```text /Applications/Google Chrome.app/Contents/MacOS/Google Chrome ``` ## Repository ```text skills/ editorial-card-screenshot/ docs/ examples/ ``` ## Docs - [`docs/agent-usage.md`](./docs/agent-usage.md) - [`docs/editorial-card-prompt.md`](./docs/editorial-card-prompt.md) - [`docs/project-overview.md`](./docs/project-overview.md) - [`docs/github-metadata.md`](./docs/github-metadata.md

话题