Workflow: Automating flows for creating literature notes

Hello there,

for a while now, I’ve been using a fairly elaborate system of automation to get the things that I read, the ideas I think of, the facts I encounter and the arguments that intrigue me into my note-taking system which centers on Obsidian. I thought it might interest you and give some pointers on how to automate some of the more tedious tasks of idea capture. The automation is mainly aimed at facilitating creating literature notes from which I then create atomic zettels.

It is based on a couple of tools chained together by another tool called IFTTT and some kind of custom API, that I hacked together running on a private server. It does little more than string manipulation, connecting and adding notes to certain files.

Online Articles

Whenever I find an interesting article or something I want to explore further, I’m adding it to my read-later tool Instapaper. There are a couple of IFTTT automations at work: For example, an article that I star in my RSS reader (Inoreader) gets automatically pushed to Instapaper. Also, if I star or heart a tweet on Twitter the first link within this tweet will also be automatically pushed to Instapaper.

While reading the article on Instapaper, I am using the highlighting function to mark the points and ideas that I want to remember and use the great connection of Instapaper to IFTTT to send the text contained in those highlights right to my API. That collects them in a temporary file so that I get all the highlights of this article in one place. I’m also able to add comments to highlights or as a kind of summary within the article. They also get sent to my API via IFTTT.
When I finish the article, I archive it in Instapaper which then (via IFTTT) triggers another event in my API that takes the temporary file with the highlights and the comments and pushes it to my vault’s inbox (via OneDrive).
There is some kind of proto syntax that I use to indicate whether a comment is just referring to a particular highlight or should be used as a kind of summary of the whole article. So when I read the article I can instantly write a small summary which is then put in a prominent position within the note file.


Another way to get information into my Obsidian system are Kindle notebooks, that is collections of highlights and comments I make while reading a Kindle book or another eBook on my Kindle reader. From the Android app you can export those notebooks. I save them to a specific folder in my OneDrive which then triggers an IFTTT action sending the file to my API. That does some parsing and transforming the note from the HTML format that Kindle outputs into a legible and usable markdown note in my vault’s inbox. This is especially helpful since I made it a habit of writing a short chapter summary after every chapter I read, especially, but not limited to, nonfiction books.

Fleeting Thoughts

My third way of getting information into the system is an IFTTT text widget which sits on the home screen of my mobile phone. It simply opens up a text window that I can use to dictate or type text into, which is then - you guessed it - sent to my API and saved in a temporary day notes file. At the end of each day another IFTTT automation uses my API to take this file and push it into my Obsidian inbox.

So much for now. if you’ve got any question or would like me to elaborate on some point, just ask :).


Thanks for sharing this very interesting process. In this post, you mentioned quite a few times regarding this API. Can you elaborate how to use that API in your system? Maybe with some technical details, e.g., which tools/services are used, if possible?

1 Like

I’d also be very interested