Atomize note

Use case or problem

To make the most of the graph tool, you are almost obliged to use a very inflexible structure: one note = one node. This is impractical if you like to take longer notes with different sections.

E.g: I want to take notes about biology. Right now I have two incomplete solutions:

  1. Create many independent notes, and add the tag neurology to all of them. This allows me to make the most of the graph, but there is no internal organization to the tags. Everything feels disorganized

  2. Create a long note with different sections, headings, etc where I write everything I want in a structured way. Right now, this is represented as a single note, which renders the graph view almost useless

Proposed solution

There is a simple solution to this. Allow the nodes in the graph to be the headings within a note. For instance: I can create a note called Biology, and it has the following sections:







We can imagine that there is an option called “Atomize note”. This function would create a node for every heading. The headings would automatically be connected in a hierarchical way. Biology would have links to evolution, physiology and taxonomy. Taxonomy would be connected to plants and animals. Etc

This solution would make Obsidian much more powerful, having the best of both worlds. You would be able to take long structured notes on a topic and at the same time see the connection among your different notes using the very useful graph view.


Take a look at …

Extract Highlights Plugin

… and its Explode Notes Mode


or perhaps …

1 Like

Wouldn’t embedding solve it?

When I have a similar cases to you I use a single folder and each section is a different note.
Where I also create a master note which embeds all other notes.
This way each node has its own freedom to set its metadata while in graph they are all connected to a single master.

This is why I find the ability to embed the most important feature of a modern PIM.


Thanks for the proposed solutions!

They are close to what I want, but not exactly

The extract highlight plugin is cool, but a bit awkward to use. If I want notes for the headings I have to highlight the headings one by one. And in the end, the solution is always to split your markdown into dozens of independent notes. Same for the refactor plugin


To make life a little easier you can open the .md file in a regular text editor & do a Find/Replace e.g. …

Find: "## "
Replace: “## ==”

EDIT: it will also need a regular expression Find/Replace to add == at the end.

… but you are still left with the other issues.

I believe this is a duplicate of Display headings and blocks in graph view.

Yes, I think you are right. That’s pretty much what I had in mind