Mastering Communication Between Shiny Modules

Preface

While working on the {kitems} package which aims at providing a framework to manage tabular data inside R Shiny apps and has its main component delivered as a Shiny module server, I had to think about several implementations for this module to be used in complex apps.

The standard and most simple one is to fully delegate the items management to the module server. In this case, all operations are performed within the module server context and the components (buttons, inputs, table) are used in the parent app.

But sometimes (actually quite often), you need more flexibility over the items or maybe you want to implement the custom ui components and this will be based on a mixed implementation where both the module server and the parent app or module will interact with the data and with each others.

In the recent months (since the beginning of 2025), this project entered a phase where I am creating apps & tools based on this still converging package to validate the core concepts through different use cases and implementations.

During this phase, one of the main challenges I faced is the communication between the module server and the rest of the app.

Some of this challenges could be solved by reading existing documentation or Q&A’s, but my overall feeling is that there’s a lack of documentation out of the basic examples in this domain and I had to spend a lot of time exploring how the communication actually works.

This guide is meant to not only be the foundation of {kitems} module server communication, but to extend beyond this framework and be used in any complex Shiny application.

I hope it will help you put all the pieces together, and help us all make smart choices.