smaug

Why we included this project

If your Twitter/X bookmarks have piled up into a backlog of half-read links, Smaug turns them into a searchable markdown library you actually own. It pulls your saved posts (bookmarks or likes), expands shortened links, and fetches the real content behind them, such as GitHub repo metadata and READMEs, external articles, and X's native long-form posts. Each item is then passed to an AI model through the Claude Code or OpenCode CLIs, which analyzes and categorizes it into folders like knowledge/tools and knowledge/articles, with metadata attached. That way researchers or anyone maintaining a personal reading list can turn a private stack of links into an organized knowledge base without depending on a third-party service. The Node.js CLI lets you set your own category rules and track token usage, so you control how the archiving is batched and how your content gets classified.