Vibe Check: Sandboxing, Skills, and Surviving Dev to Prod

​Every team building with AI agents hits the same wall. The environment works on one machine, breaks on the next, and the agent itself makes it worse because it's writing code, pulling dependencies, and touching your system in ways you didn't plan for.

​In this Vibe Check, Melanie and co-host Cecil Phillip are joined by James Bayer, CPO at Flox, to unpack what reproducibility actually means when AI is in the loop, why most setups fall apart between dev and production, and what changes when your environment, your agent harness, and your skills all come up together. They'll get into where sandboxing genuinely protects you, what observability looks like for agent code, and the tradeoffs between local and remote dev sandboxes.

​The Temporal Go getting started will be on screen as the working example, but the ideas apply to any stack where you're trying to make AI dev work survive the jump from your laptop.

​Come with questions, bring your setup, and get real answers live.