A tool designed to automatically archive your Twitter/X bookmarks and optionally likes into a structured Markdown knowledge base. It targets users seeking to preserve valuable online content, offering an AI-enhanced, organized repository of links, articles, and GitHub repositories, transforming ephemeral bookmarks into a persistent, organized knowledge library.

How It Works

Smaug uses the bird CLI to fetch Twitter/X bookmarks/likes, expands URLs, and extracts content from linked pages. It then employs Claude Code for intelligent analysis, categorization, and metadata generation, automatically filing items into distinct directories like knowledge/articles/ and knowledge/tools/, creating a persistent, organized knowledge library.

Quick Start & Requirements

Estimated setup: 5 minutes.

1. Install bird CLI: Follow instructions at https://github.com/steipete/bird.
2. Clone Smaug and install dependencies:

   git clone https://github.com/alexknowshtml/smaug
   cd smaug
   npm install
   

3. Run setup wizard: npx smaug setup (guides Twitter credentials, config).
4. Execute: npx smaug run.

Prerequisites: Node.js, Twitter session cookies (auth_token, ct0), Claude Code access. Optional: pm2 for automation.

Highlighted Details

• Automated archiving of Twitter/X bookmarks and likes to Markdown.
• Content extraction from linked articles, GitHub repos, and quote tweets.
• AI-powered categorization and filing via Claude Code into structured directories.
• Support for custom categories and mapping Twitter bookmark folders to tags.
• Automation options (PM2, cron, systemd).
• Token usage tracking and cost optimization with Claude Haiku subagents.
• Experimental media attachment handling (requires bird CLI fork).

Maintenance & Community

Maintained by alexknowshtml. Relies on bird CLI and Claude Code. No specific community channels or roadmap links are detailed in the provided README.

Licensing & Compatibility

Released under the MIT License. This permissive license allows for commercial use, modification, and distribution, making it compatible with most projects.

Limitations & Caveats

Fetching all bookmarks requires building the bird CLI from source, as the npm release lacks pagination support. Large bookmark processing can incur significant Claude API costs; use batching and cost tracking. Experimental media handling depends on a specific bird CLI fork. Requires users to manage expiring Twitter session cookies.