Automated Content Machine: From Gemini Research to a Published WeChat Article with Cover
微信公众号每日定时自动发布文章：从 Gemini 深度研究、自动markdown排版到发布带封面的微信公众号文章
Hey n8n community! 
大家好！
I’m excited to share a workflow I’ve built to completely automate my content creation pipeline. If you’re tired of the manual grind of researching, writing, creating visuals, and publishing articles, this might be the solution you’re looking for.
如果你也和我一样，厌倦了手动搜集资料、撰写文章、制作封面图、排版发布的繁琐流程，那么这个工作流可能会给你带来惊喜。
This workflow takes a single topic as input and outputs a fully-fledged, polished article published directly to a WeChat Official Account, complete with an AI-generated cover image.
你只需要提供一个主题，这个工作流就能自动完成从研究、写作、润色、配图到最终发布到微信公众号的全过程。
What This Workflow Does / 这个工作流能做什么？
- Deep Research with Gemini: It starts by taking a topic and feeding it to Google’s Gemini Pro to generate a comprehensive, well-researched draft report.
2.Gemini 深度研究：* 工作流从一个指定的主题开始，调用 Google Gemini Pro 2.5 API 进行深度研究，生成一份内容详实的初始报告。 - AI-Powered Polishing: The raw report is then sent to OpenAI’s ChatGPT (GPT-4) to refine the language, improve the structure, and ensure it reads smoothly.
4.AI 文本润色：* 将 Gemini 生成的初稿发送给 OpenAI 的 ChatGPT （谷歌的gemini也行），对文章进行语言润色、结构优化和逻辑梳理，使其更具可读性。 - Smart Content Extraction: The workflow automatically extracts a compelling Title, a concise Summary, and a descriptive Cover Image Prompt from the polished content.
6.智能内容提取：* 接着，工作流会自动从润色后的文章中，智能提取出抓人眼球的文章标题、精炼的内容摘要，以及用于生成封面图的图像提示词 (Prompt)。 - AI Cover Image Generation: The generated prompt is then fed into ChatGPT to create a unique and relevant cover image for the article.
8.AI 封面图生成：* 利用上一步生成的提示词，调用 ChatGPT 内置的 图像生成 模型，创造一张与文章内容高度相关的原创封面图。 - Automated Publishing: Finally, the workflow gathers the title, the polished content, the summary, and the new cover image, and publishes it as a new article to a specified WeChat Official Account.
10.自动化发布：* 最后，工作流会将处理好的标题、正文、摘要和封面图整合，通过接口自动发布为一篇全新的图文消息到指定的微信公众号。
Tech Stack & Key Nodes / 主要技术栈和节点
- Google Gemini Pro: For the initial research and report generation. / 用于核心内容的调研和生成。
- OpenAI (ChatGPT & DALL-E 3): For content polishing and cover image creation. / 用于文本润色和封面图创作。
- Set & Code Nodes: To manipulate data, extract specific parts of the text, and format API requests. / 用于处理数据流、提取关键信息和格式化API请求。
- HTTP Request Node: To interact with the WeChat Official Account API for publishing. / 用于和微信公众号后台API进行交互，完成素材上传和文章发布。
Potential Use Cases & Adaptations / 应用场景与改造思路
While this workflow is built for WeChat, the core logic is highly adaptable. You could easily swap the final node to publish to WordPress, Ghost, Medium, or any other platform with an API!
尽管这个工作流的最终目标是微信公众号，但它的核心逻辑完全可以复用。你只需要简单修改最后一个发布节点，就可以将文章发布到 WordPress、Ghost、Medium 或任何支持API发布的平台！
How to Set It Up / 如何配置和使用
To get this working, you’ll need:
要让这个工作流顺利运行，你需要准备以下凭证：
- Google Gemini API Key. / Google Gemini API 密钥。
- OpenAI API Key. / OpenAI API 密钥。
- WeChat Official Account Credentials: Your AppID and AppSecret with API access enabled. You might need to adjust the API calls in the HTTP Request node to match your specific WeChat setup (e.g., uploading media first to get a
media_id).
4.微信公众号开发者凭证：* 已认证公众号的开发者ID (AppID) 和开发者密码 (AppSecret)。你可能需要根据微信接口文档，微调 HTTP Request 节点（例如，需要先上传图片素材获取media_id）。
If you are interested ，contact me [email protected] !
如果感兴趣，可以和我联系[email protected]！
I hope you find this workflow useful! Feel free to ask any questions, suggest improvements, or share how you’ve adapted it for your own projects. I’d love to see what you build with it.
希望这个工作流对你有帮助！欢迎大家提问、提出改进建议，或者分享你基于它改造出的新玩法。期待看到大家的创意！
Happy automating!
祝大家自动化愉快！