Are you a compulsive collector of interesting links, articles to read later, random notes, and inspiring images? Well, good news, so am I, and I found a great tool to satisfy our acute collecting urges without drowning in chaos: Hoarder!
What is Hoarder? It’s an open-source and self-hosted app that allows you to save anything and everything in just two clicks: links with automatic previews, text notes, and even images. Okay, I know, there are already 15,000 bookmarking and note-taking apps, so why get excited about Hoarder?
Because it ticks all the boxes for the demanding accumulator:
🔍 Powerful full-text search to find that rare gem buried under tons of bookmarks in seconds.
🏷️ Automatic tagging by AI to classify and organize the mess effortlessly. Hoarder analyzes the content and assigns relevant tags. Magic!
🌙 A dark mode (essential ^^).
💾 The ability to self-host everything on your own server, to keep control over your data.
📱 Mobile apps (in addition to the web) to “hoard” from your smartphone as well.
🆓 All of this for free, with a modern look, thanks to material design.
To centralize your digital clutter, it’s ideal. You can save all the interesting articles and Twitter threads you come across, take notes when an idea crosses your mind (one day maybe? ^^), and of course store all the funniest memes to pull out at the right moment.
The creator of Hoarder, Mohamed Bassem, is a systems engineer who wanted to stay involved in web development while having fun. He used Memos, another note-taking app, but it lacked essential features like link previews and automatic tagging. Instead of complaining, he rolled up his sleeves and created his own solution. Respect.
Moreover, Hoarder is heavily inspired by Mymind, a similar commercial product, but adds the crucial self-hosting aspect for privacy paranoids like me. Open-source alternatives like Shiori or LinkWarden don’t (yet) offer automatic AI tagging.
Under the hood, it runs on sexy tech like Next.js, tRPC, Meilisearch, and OpenAI.
Installing it is very simple: everything is packaged to run easily with Docker and Docker Compose. You will also need an OpenAI API key (for tagging), but it’s optional and inexpensive. All the details are in the documentation.
You can also try an online demo if you want to get an idea before installing. Okay, I’ll leave you now, I have a bunch of links to save! 😉