I've just started work on a project that would enable GMs and players to essentially have homebrew compendium addons for D&D 5e. I haven't found a way to directly add items to an individual user's Roll20 Compendium (nor do I think Roll20 would allow such a thing) but I have discovered a possible method for allowing homebrew items to be easily made, shared, and added to a Roll20 campaign. The whole project is split up into three separate programs: Homebrew Compendium Creator (Desktop Program): This allows GMs to create their own compendium from scratch, making races, classes, spells, feats, backgrounds, monsters, and items. Their homebrew compendium can then be exported as a series of JSON files for use in the second program. Homebrew Character Creator (Desktop Program): This would be similar to the Charactermancer in that it would guide a user through the character creation process. The choices available during chargen is dependent on what's in the JSON files created by the first program. Once the user has a character made, their choices get exported as a single JSON file. Ideally, this program would also allow for a user to level up their current character, outputting a similar JSON with their additions. Character Importer (Roll20 Mod): This mod uses the JSON file created by the second program to create character sheets. It would also take JSONs created by the first program to generate monsters. Ideally, it would also take in smaller JSONs made by the first program to allow things like items or spells to be added to individual sheets, and allow for character level-ups as mentioned previously. While this seemed like the hardest step, I've thankfully already found a method that can create character sheets from JSONs. The reason the first two programs are desktop programs is because that's what I'm comfortable with. I realize the most ideal medium would be a website to increase accessibility, but I have little to no experience with web development, so I'm sticking with desktop programs to try to make development easier for myself. I've also considered trying to make them mods, too, but I feel the processes I've outlined are too UI heavy to work in a simple chat-based JS program running on Roll20. I've started scaffolding UI for the first program and have basic navigation working. My intention is to test out this entire pipeline by seeing if I can get a Feat made in program 1 onto a character sheet using the mod. That said, I'd like a bit of input on this idea before I get too deep into it: Does Roll20 own the terms Compendium and Charactermancer? Could I use these terms for these programs, or should I make more generic names for them? Would this potentially infringe on anything owned by WotC? By default, these programs would not contain any information, though in the end, I'd like to make a pre-built compendium for use in program 1 featuring everything in the SRD as a baseline. What would a homebrew compendium creator need so that every major aspect of the game could be homebrewed? So far I have the following planned: races and subraces, classes and subclasses, monsters, spells, feats, backgrounds, items (including magic items), and actions and traits shared across all monsters. I'd love any advice on making a project like this. Input, feature ideas, things to avoid- all of that would be greatly appreciated. I don't intend on stepping on any toes with this project. I know Roll20 has made an announcement about enabling custom compendiums last year, and that the task is monumental and incredibly difficult to achieve. The project I'm aiming to create here is meant to be a temporary workaround for D&D 5e until they're able to get such a feature implemented. Finally, I'd like to note that there is no guarantee that I see this project to the end. I realize the scope is massive and I've only just started, and have yet to uncover what major roadblocks may exist, if any. If this ever happens, I would try to make my progress open source to allow others to continue the work. For now, I'd like to keep it under wraps to see what everyone says about this project, and especially to see if it's viable from a software and legal standpoint.