In today’s fast-moving, AI-powered era, autonomous agents are playing a bigger role than ever. They are helping businesses run smoother and making decisions affecting millions of lives every day. While these systems are designed to make our lives easier and unlock new opportunities, we can’t get carried away—we need to implement proper AI Agent Evaluation frameworks and best practices to ensure these systems actually work as intended and follow ethical AI principles.

If I decide to add it, which solution should I pick, battle tested Sorbet or core team endorsed RBS?

What I do care about, though, is that we might start to accept and adopt opinions like “that feature is bad”, or “this sucks”, without ever pausing to question them or explore the feature ourselves.

Can I try the endpoints before I implement my application?

@adisos if reform-rails will not match, I suggest to use: https://github.com/orgsync/active_interaction I've switched to it after reform-rails as it was not fully detached from the activerecord, code is a bit hacky and complex to modify, and in overall reform not so flexible as active_interaction. It has multiple params as well: https://github.com/orgsync/active_interaction/blob/master/spec/active_interaction/modules/input_processor_spec.rb#L41

So as I see it our choices are:

React Final Forms is a great library, an enhanced version of Redux Form

We are looking to use React Final Form in our application at work. Evaluating our options and reading the documentation, we couldn't figure out how to address a specific use case.

Some are not bad, others have a bit too many quirks and probably justify a bit of ranting for having wasted part of my life

One of my favorite tools to use in doing this is the CRAAP test developed by the University of California at Chico. This method requires students to evaluate a source based on its Currency, Relevance, Authority, Accuracy, and Purpose. In fact, this method could easily be applied to “traditional” sources as well.