I have a lot of respect for permissive licenses like MIT and Apache, but I have noticed indie devs and open-core founders struggling with them recently. When you release source-available code today, you often face a tough choice. You either use a traditional open-source license and accept that your work might get scraped for AI training or wrapped into a competitor's hosted SaaS, or you keep everything closed and limit your own distribution.

There seems to be an underserved middle ground for creators who want to share their code but need practical commercial boundaries. To try and address this, I built the Prism License Framework (PLF).

PLF is a composable license generator. Instead of a monolithic approach, it lets you assemble a variant that matches your specific situation. You start with a core grant of rights and apply optional restriction modules.

For example, you can generate a source-available license that explicitly:

    Allows general commercial use

    Restricts AI dataset training

    Prohibits white-labeling

    Limits hosted SaaS wrapping

I also know that custom licenses are usually a nightmare for corporate legal teams to review. To help prevent wild license proliferation, the generator anchors around three canonical presets (Open, Balanced, and Protected). The hope is that by sticking close to these presets, enterprise adoption remains feasible without forcing lawyers to read a brand new bespoke agreement every single time.

Licensing is a complex space, and I am still figuring a lot of this out. I know there are incredibly experienced people here with strong opinions on software licensing, so I would genuinely appreciate any feedback, criticisms, or thoughts you have on the framework, the UI, or the approach itself.

Repo: https://github.com/ScientificAJ/prism-license-framework