NanmiCoder/MediaCrawler — reverse-engineered prompt

Build me a simple but usable tool for collecting public social media content and comments from major Chinese platforms like Xiaohongshu, Douyin, Kuaishou, Bilibili, Weibo, Tieba, and Zhihu.

I want it to let me log in with a real browser, preferably by scanning a QR code, and keep that login session so I do not have to sign in every time. It should support searching by keyword, pulling a specific post or video by ID, and grabbing posts from a creator’s homepage. Please include replies to comments when the platform allows it, plus an option to make a basic comment word cloud.

A small web page would be great so I can run it without memorizing terminal commands, see progress and logs, preview results, and export data in common formats like CSV, JSON, Excel, or a simple database. If proxy support is already reasonable to add, include that too.

Please keep it focused on public data and learning use only. Look up current docs online if you need to.