Linking your Images in Obsidian with Excalidraw and ExcaliBrain

Linking Excalidraw images to Obsidian notes through structured metadata turns one-off PNG pastes into searchable, navigable knowledge objects. Instead of dropping an image into a vault where it’s hard to retrieve later, the workflow embeds the drawing inside Obsidian and attaches fields like source, author, and a “parent” concept—so the image becomes part of a connected information system.

The process starts with taking a framework image from Laura Evans Hill’s posts (via a Twitter thread) and recreating it inside Excalidraw. Using Obsidian’s command palette, the creator selects an action to create a new Excalidraw drawing embedded into the active document. The drawing is set to light mode to match the bright backgrounds of the source images, and the framework image is pasted into the canvas. From there, the key move is metadata: the creator adds a “source” field containing the Twitter link, plus tags/fields for export behavior (explicitly setting export dark mode to false), the author name, and a parent thought labeled “visual vocabulary.” A short summary can also be placed above the “text elements” section, with the important detail that anything above that section won’t be overwritten by Excalidraw, while content below it can be replaced.

To make the drawing even easier to scan and retrieve, additional items (including icon-based elements) are inserted into the metadata block. The transcript notes a practical snag: some pasted icons appear larger than others, but the size can be adjusted by adding a pipe character and specifying a pixel value (the example uses 50). Once the metadata is in place, Obsidian search becomes the retrieval engine. Searching for a term like “actionable steps” surfaces the document containing the embedded drawing, and opening it reveals the same linked image object.

The workflow also leverages ExcaliBrain to connect the drawing to the broader note graph. After embedding the Excalidraw content, the “nine abstract frameworks to visualize ideas” item becomes linked to the “visual vocabulary” note. From that parent note, the user can navigate to the related visual items and then jump back out to Laura Evans Hill’s page, including the conference session “Atomic visuals to organize your ideas” and its date. The result is a system where the image isn’t just stored—it’s indexed, categorized, and cross-referenced.

Overall, the approach trades a few minutes of upfront metadata entry for long-term payoff: faster discovery, cleaner navigation, and higher confidence that visual artifacts can be found later when they’re needed for thinking, writing, or planning. The transcript closes by encouraging viewers to join the upcoming conference session if they’re watching before Friday, framing the material as both inspiring and practically useful.