Over the last several years, I've tried all sorts of different read-it-later apps. As the name suggests, these apps are for saving articles to be read later. They usually offer a dedicated reading experience including notes, annotations, syncing with knowledge bases and so on.
The problem is that these apps create an unnecessary separation between the content we read and our knowledge bases1. It doesn't have to be like this. They can be part of the same thing. And when we remove this unnecessary separation some really cool possibilities open up.
This piece details how I've crafted such a unified system in Obsidian, to the point where I happily describe it as my ideal reading environment.
The Obsidian Web Clipper
The system begins with the obsidian web clipper browser extension. Whether I'm on my laptop or my phone, whenever I come across an article that seems worth reading I clip it directly into Obsidian. The clipper saves the full contents of the article as a markdown note in my vault.
This is what the web clipper extension looks like
The clipper lets you configure templates which can pre-fill properties on the note in a consistent format.
This is the web clipper's settings page where you can configure templates to pre-fill when clipping
What the clipped article looks like in Obsidian
These auto-filled properties can be used to create an Obsidian base with multiple views. I use this base to help me keep track of my clipped articles and easily return to them for reading. The two main views I have are a kanban view, to triage articles for reading and track their status, and a table view, to easily view my clipped articles by recency.
The kanban view of my articles base
The table view on my articles base
Not only are these articles available for reading offline and synced across all of my devices, I also now have a full-content archive that I can keep no matter what happens to the original website.
Note-Taking for Durable Understanding
As I've written about previously, I shifted most of my reading-based note-taking to flashcard writing. To me, writing and reviewing flashcards is the essence of knowledge acquisition and comprehensive learning2. These are the very knowledge exercises and reps that develop my long-term understanding.
When I write flashcards from a piece of content, I want to include context about where the flashcard comes from. That way, when I review the flashcard in the future, I have the context I need. Typically, the shape of my flashcards are like this: "According to [author] in [title of article], [flashcard question]". The first part of the flashcard is a template which can be auto-filled given the right system. This is possible to do in Obsidian, as I'll show below.
I use Anki as my spaced-repetition flashcard app of choice. Anki has a plugin that allows for the programmatic creation of flashcards, called AnkiConnect. I then use an Obsidian plugin that can sync flashcards written in Obsidian to Anki, called Obsidian_to_Anki. The plugin lets you define custom flashcard syntax, so you can write flashcards however you want. The basic idea is that it scans the vault and looks for {frontside}{delimiter}{backside}. I use ?:: as my flashcard delimiter but it can be whatever you want.
While reading the article, I simply add flashcards as notes throughout the article. It looks like this:
When I clip the article, I include a templated content prefix which includes info about the type of Anki flashcards to create and info about the article to append at the beginning of the flashcard (this is particular to the Obsidian_To_Anki plugin I use). These values can then be used by the plugin when creating the Anki flashcards, which I use to create the flashcard context prefix.
After I finish reading the article3, I activate the anki plugin and it scans my vault for any newly created flashcards.
The plugin goes through and marks them as synced so they don't get duplicated in future scans. Thats what the <!--ID: {number}--> is.
The Anki flashcard includes all the necessary context about the article (author, title, date), along with a link to the original article and a deep link to the note in Obsidian, so I can always navigate to the source directly from the flashcard if I want.
Reading E-Books
This flashcards-as-notes system works just as well for books that I read on my Kindle. Instead of reading and writing notes directly in Obsidian, I write the flashcards as notes on the Kindle. Then I use a different plugin, Kindle Local Sync, to sync these notes into Obsidian, where I then use the same Obsidian_To_Anki plugin to sync them to Anki.
Whenever I'm ready to convert my Kindle notes to anki flashcards, I'll plug my Kindle into my laptop and run the Kindle Local Sync plugin, which adds all the notes (and highlights) to a note for that book. Then I use the same Obsidian_To_Anki plugin to sync these flashcard notes to Anki.
Sometimes typos make it in from my Kindle notes. I often don't bother to fix them if it doesn't affect the legibility.
AI & Malleable Software
Obsidian has many community plugins to pick from that cover all sorts of niche requirements. Even so, you may find yourself with particular needs not well covered by existing plugins. Luckily, that doesn't need to be as much of an obstacle anymore. Since everything in your Obsidian vault (including the plugins you've installed) are just files in a directory, it is relatively straightforward to work with coding agents to make tweaks to existing plugins, create altogether new ones, or build custom software for specific workflows.
For example, the Obsidian_To_Anki plugin was difficult to configure specifically for how I wanted it to work. No worries. I can always open my vault in VS Code, activated Claude Code in the terminal, point it at the plugin files and start making the changes I want. Or I can just ask questions about how to do certain things.
The point is that we can use coding agents to shape our knowledge environments how we like. Obsidian provides a solid foundation for this, but the key is that it's all "just" local files on my computer.
Keeping It Fun
My philosophy is that learning should always be fun and rewarding. In other words, it should be intrinsically motivating. I think a big part of why so many fail to maintain flashcard reviewing as a daily habit is because it can feel more like a chore than an engaging and rewarding learning activity.
Part of the solution is having well-written, atomic flashcards that are quick and easy to review. This is a skill to develop and takes practice. The other part is quality control. If a flashcard isn't interesting to review anymore, delete it. I only want to review things that are still relevant and interesting.
This all results in a daily habit of reviewing my flashcards which usually takes between 15-60 minutes (including all my language learning flashcards). I almost always enjoy this time, since I'm reinforcing information that I intrinsically care about. The only time flashcard review feels like a chore is when I've missed a few days in a row and the backlog piles up.
What I love about this reading system is that there is no "I'll do something with these insights later" mentality, as is often the case with reading notes. Instead, every time I review a flashcard I made while reading, it becomes "I'm doing something with these insights right now".
Happy learning!
P.S. If you'd like to setup this kind of reading system yourself, let me know and I can share more info about how it works under the hood.