# Tesserae makes it actually easy to run self-hosted e-ink > Source: > Published: 2026-08-11 19:14:26+00:00 # Tesserae makes it actually easy to run self-hosted e-ink [MCP](/en/tags/mcp/)to generate highly detailed layouts automatically. The whole system is built on a "drop-a-folder" architecture, which is a massive win for extensibility. If you want to add a new widget, a custom theme, or a specific device renderer, you just drop the corresponding folder into the Tesserae directory. It doesn't require a complex build process or deep diving into the core source code just to change a visual element. In terms of out-of-the-box utility, it comes with over 30 pre-installed widgets. There is also a community marketplace for more, though the creator admits they've built most of those themselves so far. The hardware support is surprisingly broad for a project of this scale, currently supporting over 10 different panels and devices. For anyone looking for a practical tutorial on how to get started, the deployment is straightforward since it's designed to be self-hosted. Because it's licensed under AGPL, it's fully open-source and has a growing ecosystem of contributors providing firmware updates and even a community-made iOS app. If you're planning a deep dive into e-ink home automation, here is the current breakdown of the feature set: **Layout Engines:** Card-based grid system and a free-form canvas.**Integration:** MCP support for automated canvas editing.**Extensibility:** Folder-based plugin system for widgets and themes.**Hardware:** Support for 10+ different e-ink panels.**License:** AGPL open-source. The project started as a way to utilize a 13.3” Spectra 6 display, but it has evolved into a legitimate framework for anyone who wants a low-power, high-visibility dashboard without the friction of writing custom Python scripts for every single screen update. It's still got some rough edges, but for a self-hosted solution, it's significantly more polished than the fragmented scripts usually found on GitHub. It's a solid example of how a specific hardware pain point can be solved with a modular AI workflow. [Next Go is arguably the most AI-friendly language for modern backend →](/en/news/5940/)