Description

将任意内容（URL/文件/文本）生成单页幻灯片简报，输出为图片/PDF/网页。基于 frontend-slides 构建。触发条件：(1) "生成简报"、"AI简报"、"摘要" (2) 上传文件并要求生成简报

README (SKILL.md)

Briefing Generator

Name: Briefing-pro
Author: nonlinearhuman

将任意内容生成为精美的单页幻灯片简报，输出为图片（默认）/PDF/网页。

支持的输入格式

类型	处理方式
URL 链接	web_fetch 获取网页内容
Word (.docx)	python-docx 提取文字
Excel (.xlsx)	openpyxl 提取数据
PDF (.pdf)	pdfplumber 提取文字
纯文本	直接使用内容
图片	图像识别提取信息

支持的输出格式

格式	说明	适用场景
图片 (PNG)	默认输出	分享、微信
PDF	打印为 PDF	文档存档
网页 (HTML)	交互式幻灯片	演示播放

支持的风格

使用 frontend-slides 的所有预设风格：

风格名称	风格描述
Bold Signal	活力橙红，卡片式设计
Electric Studio	清新蓝绿，渐变背景
Creative Voltage	活力复古，几何图形
Dark Botanical	优雅深色，植物装饰
Notebook Tabs	杂志风格，彩色标签
Pastel Geometry	柔和粉彩，现代简约
Split Pastel	明亮分割，活泼现代
Vintage Editorial	复古编辑，幽默个性
Neon Cyber	霓虹赛博，网格背景
Terminal Green	终端绿，开发者风格
Swiss Modern	瑞士现代，极简精确
Paper & Ink	纸墨风格，文学质感

工作流程

Step 1: 识别输入类型

根据用户输入判断类型：

以 http:// 或 https:// 开头 → URL
以 .docx、.xlsx、.pdf 结尾 → 文件
其他 → 纯文本

Step 2: 提取内容

URL：

web_fetch({
  url: "链接",
  maxChars: 10000
})

Word/Excel/PDF：使用 Python 库提取

图片：使用 image 工具分析

Step 3: 提取关键信息

用规则从内容中提取：

标题：从文件名、\x3Ctitle>、第一行标题提取
数据：提取数字、百分比、关键词
要点：从段落中筛选重要句子

Step 4: 选择风格

Question: 简报风格

Header: "风格"
Question: "想要什么风格？"
Options（列出 frontend-slides 的 12 种风格）

Step 5: 选择输出格式

Question: 输出格式

Header: "输出"
Question: "想要什么格式？"
Options:
- "图片 (PNG)" - 默认，分享方便
- "PDF" - 适合打印存档
- "网页 (HTML)" - 交互式演示

Step 6: 选择页面尺寸

Question: 页面尺寸

Header: "尺寸"
Options:
- "A4" - 竖版文档
- "16:9" - 宽屏演示

Step 7: 生成并输出

根据输出格式选择：

图片：浏览器截图
PDF：浏览器打印
网页：直接打开 HTML

速度优化

跳过 AI 总结，规则提取关键信息
内容长度限制 10000 字符
复用已有浏览器窗口

错误处理

无法提取内容时，询问用户手动输入关键信息
文件格式不支持时，提示用户转换为其他格式
输出失败时，提供 HTML 文件作为备选

致谢

本 Skill 使用了 frontend-slides 进行幻灯片渲染，特此致谢。

Usage Guidance

This skill appears to do what it says (create single‑page slide summaries) and includes ready templates, but there are important safety gaps you should consider before installing or using it: 1) The runtime doc doesn't mention sanitizing or escaping extracted/remote content before embedding it into HTML templates — opening or sharing raw generated HTML could execute embedded scripts (XSS). Prefer PDF/PNG outputs or ensure the agent sanitizes HTML. 2) The SKILL.md references tools (web_fetch, python libraries, image analysis, headless browser) but doesn't declare them as required; verify those tools exist in your environment and that fetches are limited (avoid fetching private/internal URLs). 3) Templates load Google fonts which causes external network requests; if you need air‑gapped operation, adjust templates. 4) Avoid uploading highly sensitive documents unless you trust how the agent extracts and stores the content. If you want to proceed, ask the maintainer how input is sanitized, whether web_fetch is domain‑restricted, and which runtime tools/libraries are used and available — or request an explicit sanitization step in the SKILL.md.

Capability Analysis

Type: OpenClaw Skill Name: briefing-jianbao Version: 1.0.1 The skill bundle is designed to generate slide briefings from URLs and documents, but it contains multiple XSS vulnerabilities in its HTML templates (templates/professional.html, templates/vibrant.html, and others). Specifically, content extracted from untrusted sources is injected directly into <script> blocks and the DOM via innerHTML without sanitization. While these appear to be unintentional implementation flaws rather than intentional malware, they pose a risk of script execution within the agent's browser environment when rendering screenshots of malicious or adversarial input content.

Capability Assessment

✓ Purpose & Capability

Name/description (generate single‑page slide briefings) align with included HTML templates and the stated extraction methods (web pages, .docx, .pdf, images). The declared capabilities are coherent with the required libraries mentioned in the docs (python-docx, openpyxl, pdfplumber) and the frontend-slides renderer.

⚠ Instruction Scope

The SKILL.md instructs the agent to fetch arbitrary URLs (web_fetch) and to inject extracted content directly into HTML templates that contain inline JS templating placeholders. There is no mention of sanitizing or escaping user/remote content before embedding into HTML, which risks XSS if the generated HTML is opened/shared. Instructions also reference tools (web_fetch, image tool, headless browser screenshot/print) but do not constrain or detail safe handling (whitelists, size limits beyond maxChars, or sanitization steps).

✓ Install Mechanism

Instruction-only skill with no install spec and no downloaded code. This minimizes installation risk — nothing is written to disk during install by the skill bundle itself.

ℹ Credentials

The skill requests no environment variables or credentials (good). SKILL.md expects Python libraries and tooling (python-docx, openpyxl, pdfplumber, image tool, browser rendering) but does not declare them as required; this is not dangerous per se but is an omission the integrator should verify. Templates load fonts from Google (external network requests).

✓ Persistence & Privilege

always:false and no special persistence or configuration writes are requested. The skill is user-invocable and may be invoked autonomously (platform default) but there's no elevated privilege requested by the skill itself.

Version History

v1.0.1

🎯 Briefing Pro AI-Powered One-Page Slide Generator Turn any content (URL, file, or text) into a stunning one-page slide deck—output as image, PDF, or HTML. Supported Input Type How It's Processed URL web_fetch to extract content Word (.docx) python-docx for text Excel (.xlsx) openpyxl for data PDF pdfplumber for text Plain Text Direct use Image AI image analysis Supported Output Format Description Best For Image (PNG) Default Sharing, WeChat PDF Print-ready Archiving HTML Interactive slides Presentations 12 Built-in Styles Style Vibe Bold Signal Energetic orange, card-based Electric Studio Fresh blue-green, gradient Creative Voltage Retro-vintage, geometric Dark Botanical Elegant dark, botanical Notebook Tabs Magazine style, colored tabs Pastel Geometry Soft pastel, modern minimal Split Pastel Bright split, playful Vintage Editorial Retro editorial, witty Neon Cyber Neon cyberpunk, grid Terminal Green Dev-style terminal Swiss Modern Minimalist precision Paper & Ink Literary, paper texture Triggers "Generate briefing" / "AI briefing" / "Summary" Upload a file and ask for a briefing Workflow Detect input type → URL / File / Text Extract content → Fetch & parse Extract key info → Title, data, key points Select style → Choose from 12 presets Select output → Image / PDF / HTML Select size → A4 portrait / 16:9 widescreen Generate & deliver → Browser screenshot / print / HTML Why Briefing Pro ✅ Fast — Rule-based extraction, no AI delay ✅ Beautiful — 12 professionally designed styles ✅ Flexible — Multiple input/output formats ✅ Portable — Perfect for sharing Best for: Quick summaries, shareable insights, presentation slides.

v1.0.0

一句话介绍：把任意内容（网页/文件/图片）一键生成精美幻灯片简报，支持图片、PDF、网页三种格式输出。基于 frontend-slides 构建。核心功能： 🌐 支持 URL、Word、Excel、PDF、图片、纯文本多种输入 🎨 12 种预设风格可选（赛博朋克、瑞士极简、复古杂志...） 📱 输出 PNG 图片（方便分享微信）、PDF（打印存档）、HTML（演示播放） ⚡ 纯规则提取，无需 AI 加速生成使用场景：老板要你 "把这份报告整理成 PPT" → 甩个链接就搞定读了一篇好文章 → 一键生成要点简报群里有价值的讨论 → 截图识别自动总结适用人群：需要频繁做汇报、写总结、做分享的职场人

Metadata

Slug briefing-jianbao

Version 1.0.1

License MIT-0

All-time Installs 1

Active Installs 1

Total Versions 2

Frequently Asked Questions

What is Briefing-pro?

将任意内容（URL/文件/文本）生成单页幻灯片简报，输出为图片/PDF/网页。基于 frontend-slides 构建。触发条件：(1) "生成简报"、"AI简报"、"摘要" (2) 上传文件并要求生成简报. It is an AI Agent Skill for Claude Code / OpenClaw, with 322 downloads so far.

How do I install Briefing-pro?

Run "/install briefing-jianbao" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Briefing-pro free?

Yes, Briefing-pro is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Briefing-pro support?

Briefing-pro is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Briefing-pro?

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

More Skills

Briefing-pro