6 articles
Cursor's @-mention works on any text file in your repo. The bottleneck for AI-coding workflows isn't Cursor's intelligence — it's getting outside research material into a format Cursor can @ reference. Here's the workflow that 10x'd my Cursor output.
Reddit blocks every server-side scraper — Firecrawl, Jina, even Reddit's official API for unauthenticated reads. The workaround: read the thread from inside your own logged-in browser. Here's the full workflow for getting Reddit content into Claude as context.
Cursor and Cowork need context to be useful. The best context source you already have is the Claude conversation where you actually figured the problem out — but it's stuck behind a chat UI. Here's how to get it out.
WeChat Official Account articles carry some of the highest-signal long-form Chinese content on the internet. They are also the hardest to scrape — Tencent blocks every external HTTP request. The reliable path in 2026 is a browser extension that reads the article from the browser tab you already opened it in.
Xiaohongshu blocks every server-side scraper — the anti-bot signing rotates monthly and pure HTTP requests get nothing back. Here is the only reliable way to get Xiaohongshu post content into Markdown for Claude, ChatGPT, or your own RAG pipeline: read the page from the browser you already opened it in.
Markdown is quietly becoming the standard input format for AI workflows. Here are five practical ways to use it — from RAG pipelines to fine-tuning datasets — with concrete examples and tooling recommendations.