为自然搜索排名提供站点审计、内容撰写与竞品分析。
设计与多媒体
LinkedIn Poster
试用Use when you need to post content to LinkedIn, including text posts, image posts, and
它能做什么
Post text and images to LinkedIn via the tool using the linkedin channel. Supports text-only posts, image posts (local files or URLs), and #BuildInPublic updates.
技能文档
LinkedIn Posting
Overview
Post text and images to LinkedIn via the message tool using the linkedin channel. Supports text-only posts, image posts (local files or URLs), and #BuildInPublic updates.
Prerequisites
The linkedin channel must be configured in channels.linkedin with:
accessToken: OAuth 2.0 token withw_member_socialscopepersonUrn: Your LinkedIn person URN (e.g.urn:li:person:Abc123)
Posting Text
Use the message tool to send a text post:
message send --channel linkedin --text "Your post content here"
The post goes to the authenticated user's LinkedIn feed as a public post.
Posting with Images
Two-step process:
- Create the image — generate or locate the image file (PNG/JPG/GIF/WebP)
- Post with image — use
message sendwith the--mediaflag:
message send --channel linkedin --text "Post caption" --media /path/to/image.png
Supported image sources:
- Local file paths:
/Users/david/image.png file://URLs:file:///Users/david/image.png- Remote URLs:
https://example.com/image.png
Image Upload Flow (internals)
The extension handles a 3-step LinkedIn UGC API flow automatically:
- Register upload → gets
uploadUrl+assetURN - PUT binary to
uploadUrl - Create UGC post with
assetreference
You don't need to manage this — just provide the image path.
Best Practices
- Always confirm post content with the owner before posting
- Keep posts concise — LinkedIn favors 150-300 word posts
- Add hashtags at the end, not inline (3-5 max)
- For #BuildInPublic: include what you built, what you learned, a visual
- Image sizing: 1200x627px for link previews, 1080x1080 for square posts
- Never post without explicit owner approval
Post Templates
#BuildInPublic Update
🚀 [What you built today]
[2-3 sentences about what it does and why it matters]
Tech stack: [list]
[Screenshot or diagram]
#BuildInPublic #[topic] #[tech]
Project Milestone
✅ [Milestone achieved]
[Brief context — what, why, how]
Key learnings:
→ [Learning 1]
→ [Learning 2]
→ [Learning 3]
Next up: [what's coming]
#[project] #[topic]
Technical Deep-Dive
💡 TIL: [Thing you learned]
[Explanation in 2-3 paragraphs]
[Code snippet or diagram if relevant]
#TIL #[topic] #[tech]
Troubleshooting
| Error | Cause | Fix |
|---|---|---|
| 401 Unauthorized | Token expired | Re-authenticate, update channels.linkedin.accessToken |
| 403 Forbidden | Missing w_member_social scope | Re-create OAuth app with correct scopes |
| Image upload 400 | File too large (>10MB) | Compress image before uploading |
| Empty postUrn | API didn't return ID | Post likely succeeded — check LinkedIn feed |
相关技能
以 AI 机器人身份加入视频会议,提供语音、虚拟形象与屏幕共享四种模式。
docx
官方通过 docx-js 脚本、XML 补丁和 pandoc 解析来创建、读取和修改 Word .docx/.dotx 文件。
用 Python 处理常见 PDF 任务:合并、拆分、抽取文本与表格、旋转页面、OCR 扫描件、生成新 PDF。
用文档驱动的 MCP 服务器与小组件脚手架,搭建 ChatGPT Apps SDK 应用。
用已发布的设计系统组件拼出整页 Figma 屏幕,而不是手绘基础元素。