Siftly is a self-hosted Twitter/X bookmark manager that organizes and visualizes your bookmarks locally using AI categorization.
The tool addresses the problem of bookmark sprawl by running a four-stage AI pipeline on imported bookmarks to automatically categorize them, extract key information, and make them searchable by meaning rather than keywords alone. Everything runs on your machine using SQLite for storage, with only the AI API calls going to external services. The application provides semantic search across bookmarks, an interactive mindmap visualization of your bookmark graph, filtered browsing in grid or list views, and export capabilities for media and data in multiple formats.
Siftly suits developers and knowledge workers who want to retain full control over their bookmark data without relying on cloud services or browser extensions. The project is particularly valuable for those already using Claude through the Code CLI, since it can automatically detect and use that authentication with zero configuration. For others, it supports API key authentication through the settings interface, environment variables, or compatible Anthropic endpoints. The tool is built with Next.js and TypeScript, making it accessible to developers comfortable with modern web stacks who want to self-host their bookmark infrastructure.
The project shows active development with regular commits addressing both features and bug fixes. Work spans across the full stack including database schema improvements, UI refinements, and pipeline enhancements. The codebase demonstrates attention to user experience through thoughtful authentication detection and streamlined setup flows. Development activity indicates ongoing maintenance of core functionality alongside incremental feature additions.
