Try stimulus with Jekyll
Evolution Note: This article was written in March 2018 and reviewed in October 2026. While my day-to-day engineering now centers on distributed automation backends, I still appreciate static site pipelines that prioritize immediate page loads. For recent technical writings, head over to the homepage.
Static site generators like Jekyll produce fast, cacheable HTML without server-side rendering delays. Even so, editorial blogs and documentation portals occasionally require interactive UI details, such as collapsible accordions or dynamic search filters.
Reaching for a full single-page framework often brings heavy bundle pipelines and hydration delays. Stimulus, created by Basecamp as part of Hotwire, takes a different route by augmenting existing markup without taking over page rendering.
Progressive Enhancement with Stimulus
Stimulus connects JavaScript behavior directly to existing HTML tags using custom data attributes:
The DOM remains the single source of truth. Rather than generating templates in memory on the client, controllers bind to elements using attributes like data-controller, data-action, and data-target.
Event handling also connects automatically. Whenever matching elements appear or leave the document, the controller lifecycle hooks fire without requiring manual event listener cleanups in your script files.
Integrating with a Jekyll Pipeline
In a standard Jekyll site, adding Stimulus involves three simple steps:
- Serve a small bundled JavaScript file containing your controller classes via the assets directory.
- Add the corresponding
data-controllerattributes into your Liquid post layouts. - Keep the rendered HTML cacheable so search bots receive the full text immediately on the first network round trip.
For a minimal working setup, check out the sample repository: jekyll-stimulus on GitLab.
Practical Takeaway
Web development regularly circles back to server-rendered foundations. Pairing static generators with small micro-frameworks like Stimulus or Alpine gives you interactive components while preserving fast initial paint times and search accessibility.
Working on complex AI or automation pipelines?
Let's discuss system architecture, deterministic state boundaries, or workflow design.
Grab a Coffee