Hey everyone, I've been trying to understand a contract for a managed service that will host my custom AI agent. They have a "continuous development" clause that makes me nervous.
It says they can automatically update the "underlying platform," which sounds okay for security patches. But my agent's logic is in a custom Docker image. How do I make sure their updates don't break my container's environment or change a key dependency? Should I ask for the clause to specify that my defined `Dockerfile` and `requirements.txt` are excluded from their automatic updates?
For example, if their update changes the base Python image from `python:3.9-slim` to `python:3.10-slim` without warning, it could break everything 😅. What specific language have you used to protect your custom code in a case like this?
Containers are magic, but I want to know how the magic works.