Everywhere (at conferences, in various GH repos) I see people completely freely sharing their ways of using AI. Employees of several companies I've recently passed through without any second thought put their guardrails into shared repos, prompts for how to do code self-review. Hell: manual testers are currently sitting and writing an entire agent that will replace them once completed.
For me this is shooting yourself in the foot: the prompts I've developed over the last dozen or so months are one of the few things that make me even slightly irreplaceable.
Am I thinking correctly, or am I overreacting?