What is mfe-canvas?

The Micro-Frontend Canvas is a one-page structured document template for software teams who split their web applications into independently developed pieces, a pattern called micro-frontends. Each piece of the application is its own small project owned by a separate team, and the canvas is a worksheet for thinking through what that piece does, where its edges are, how it talks to other pieces, who owns it, and what risks it carries. The idea is to fill out one canvas per micro-frontend, either before building it to force the team to make decisions upfront, or after the fact to audit whether the boundaries and ownership actually match what was planned. The README describes two main workflows: a half-day workshop where the team prints the blank canvas and works through it together, or an async approach where a tech lead fills it to 90 percent and the team reviews it together afterward. The repository itself is mostly documentation rather than code. It contains a PDF guide with a blank template and worked examples, a Markdown version of the template you can copy directly into your project repository, and several filled-in example canvases covering an e-commerce checkout flow and an e-learning video player. Community-contributed examples are also welcome. The README distinguishes the canvas from a regular README file: the canvas captures why a micro-frontend exists and what decisions were made about its boundaries, while the README captures how to set it up and work on it day to day. The canvas changes infrequently, the README changes often. Both live in the repository together. There is also a section on using the completed canvas as context for AI coding assistants, where providing structured boundary and constraint information is described as improving code generation quality compared to vague prompts. The author is Luca Mezzalira, who wrote the O'Reilly book on micro-frontends. The template is licensed for non-commercial use under Creative Commons.