Thanks everyone for the thoughtful input — lots of good perspectives here.
This is exactly the kind of discussion I was hoping to spark.
Just to clarify the angle I was trying to explore: this isn’t only about “AI inside Xojo”.
The core point is how Xojo can participate in the newer agent-centric workflows that are already emerging: a loop where the developer, autonomous agents and sub-agents (often running outside the IDE) all collaborate on the same project.
In that context, the key requirement is staying in the loop — allowing external agents to read, reason about, and safely modify a Xojo project, while the IDE remains a first-class part of the cycle rather than the only place where work happens.
Thanks again for the thoughtful discussion - curious to see how this evolves.