Just started kicking the tires on Amazon Q Developer this week, and I'm genuinely impressed with the code suggestions and the CLI integration. It's snappy! But while going through the setup, I hit the documentation about data handling, and now I can't unsee it.
The policy states that chat logs and inputs may be retained for up to 30 days to "improve services." I work in a heavily regulated industry, and even though I'm using it for personal open-source projects, the idea of my code context, architectural questions, and even snippets of proprietary libraries from my dependencies sitting in a log somewhere for a month gives me pause.
* Does this mean prompts containing internal module names or code structures are stored?
* Is there a way to opt-out or auto-purge chats on a shorter cycle?
* How does this compare to other AI coding assistants you've used?
I want to integrate this into my CI/CD and local automation flows, but data governance is a non-negotiable first step. It feels like a major point that's getting lost in the feature hype. What's your read on this? Are you just avoiding putting any sensitive info in, or is there a workaround I've missed?
Ship fast, measure faster.