技能详情（站内镜像，无评论）
许可证：MIT-0
MIT-0 ·免费使用、修改和重新分发。无需归因。
版本：v1.0.0
统计：⭐ 0 · 24 · 0 current installs · 0 all-time installs
⭐ 0
安装量（当前） 0
🛡 VirusTotal ：可疑 · OpenClaw ：可疑
Package：aowind/sjht-ppt-delivery
安全扫描（ClawHub）
- VirusTotal ：可疑
- OpenClaw ：可疑
OpenClaw 评估
The skill's code mostly matches its stated purpose (convert slides to PDF and send via Feishu) but it silently reads local OpenClaw config for credentials and the registry metadata did not declare that or any required credentials — an incoherence you should review before installing.
目的
The skill's name/description (HTML→PDF + Feishu delivery) matches the included scripts: html2pdf.cjs produces screenshots->PDF and send_file_feishu.py uploads via Feishu API. However, the registry metadata declares no required env vars or config paths, while the runtime actually depends on Chromium, global npm packages (puppeteer-core, pdf-lib) and reading ~/.openclaw/openclaw.json for Feishu credentials. That undeclared dependency is a mismatch.
说明范围
SKILL.md instructs running the included scripts and to pull the recipient id from inbound metadata — that is reasonable. But the instructions also assume an OpenClaw config file (openclaw.json) exists with Feishu APP_ID/APP_SECRET and do not document that this file will be read. The runtime will load the entire config file from the user's home directory (~/ .openclaw/openclaw.json), which is broader access than the skill's metadata claims.
安装机制
No install spec (instruction-only + included scripts). The scripts rely on locally installed Chromium and global npm packages (puppeteer-core, pdf-lib) and Python requests — these are plausible for the task. No external downloads or archive extraction are performed by the skill bundle itself.
证书
The skill did not declare required env vars or config paths, yet send_file_feishu.py reads ~/.openclaw/openclaw.json (and will fall back to FEISHU_APP_ID / FEISHU_APP_SECRET env vars if present). Reading a full config file can expose unrelated secrets stored there. Requesting Feishu credentials is reasonable for delivering files, but the skill's metadata should have declared that it needs access to that config path or to explicit env vars.
持久
The skill does not request always:true, does not modify other skills or system configuration, and does not persist new credentials. It runs on demand and has no elevated persistence behavior.
安装（复制给龙虾 AI）
将下方整段复制到龙虾中文库对话中，由龙虾按 SKILL.md 完成安装。
请把本段交给龙虾中文库（龙虾 AI）执行：为本机安装 OpenClaw 技能「sjht-ppt-delivery」。简介：Convert HTML slide deck to PDF and send to Feishu user. Use when the user asks …。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装：https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/aowind/sjht-ppt-delivery/SKILL.md
（来源：yingzhi8.cn 技能库）SKILL.md
---
name: ppt-delivery
description: |
Convert HTML slide deck to PDF and send to Feishu user. Use when the user asks to
generate a PPT/presentation and deliver it as a PDF file via Feishu message. Triggers
on: "做PPT", "做个演示", "生成PPT", "做幻灯片", "发送PDF", "转PDF", "做报告",
or when a PPT/slide task is completed and needs delivery. This skill covers the full
pipeline: HTML slides → font scaling → PDF conversion → Feishu file upload & send.
---
# PPT Delivery — HTML 演示文稿转 PDF 并发送飞书
完整流程：生成 HTML 幻灯片 → 放大字体 → 转 PDF → 上传飞书发送给用户。
## 前置依赖
- `chromium-browser`（已安装）
- `puppeteer-core`（全局 npm 包）
- `pdf-lib`（全局 npm 包）
- Python 3 + `requests`（已安装）
- 飞书机器人已配置（openclaw.json 中有 APP_ID/SECRET）
## 工作流程
### Step 1: 生成 HTML 幻灯片
使用 `frontend-slides` 或 `jobs-style-ppt-generator` skill 生成 HTML 文件。
### Step 2: 字体放大（必须）
用户通常反馈字体太小，默认执行两轮放大：
**第一轮放大**（CSS 修改）：
- body font-size → `22px`
- 所有 ≤1rem → ×1.25
- 所有 1~1.5rem → ×1.35
- 所有 clamp() 值 → ×1.2
- 卡片 padding → ×1.2
**第二轮放大**（如用户仍嫌小）：
- body font-size → `26px`
- 所有字号 → 再 ×1.2
- clamp() 值 → 再 ×1.15
- 卡片 padding → 再 ×1.15
始终保持标题/正文层级关系。
### Step 3: HTML 转 PDF
使用脚本逐 slide 截图嵌入 PDF，保证视觉一致性：
```bash
NODE_PATH=$(npm root -g) node <skill_dir>/scripts/html2pdf.cjs <input.html> <output.pdf>
```
参数：
- `--width 1920`（默认）
- `--height 1080`（默认）
输出：多页 PDF（每页一张幻灯片截图）。
### Step 4: 发送飞书文件
将 PDF 通过飞书 Bot API 发送给用户：
```bash
python3 <skill_dir>/scripts/send_file_feishu.py <pdf_path> <user_open_id>
```
user_open_id 从消息的 inbound metadata `sender_id` 获取。
## 完整示例
```
1. UI agent 生成 /root/projects/report.html
2. 字体放大（两轮）
3. NODE_PATH=$(npm root -g) node ppt-delivery/scripts/html2pdf.cjs /root/projects/report.html /root/projects/report.pdf
4. python3 ppt-delivery/scripts/send_file_feishu.py /root/projects/report.pdf ou_xxxxx
5. 回复用户："PDF 已发送 📎"
```
## 注意事项
- PDF 文件大小通常 1-3MB（5-10 页）
- 如果 chromium 截图有渲染问题，检查字体是否加载完成（脚本内置 3 秒等待）
- 飞书发送需要 bot 有 `im:message:send_as_bot` 权限
- 文件类型支持：pdf、doc、xls、ppt、mp4、opus