That’s the thing: if you review the agents’ output yourself, you won’t really keep the benefits of using them.
And this is something many senior engineers are struggling with. I believe the value is in driving agents with proper guardrails and quality checks and go YOLO mode.
I’m significantly older than you. I started coding in the late 60s. My current strategy is to not read any of the code written by my agents. That’s the only way I can take advantage of their productivity. What I do instead is to surround the agents with extreme constraints. Unit
@thsottiaux Codex feature request for better UX:
If the user hasn't updated Codex in a while and has accumulated multiple pending updates, allow Codex to install them all in a chain of updates by clicking the "Update" button only once.
Or, if you want to preserve the current
Very promising @alex_whedon!
I’m building an agent that requires low latency and cost, so I will try SubQ for sure. Hope it’s good at tool calling too 🤞
Already signed up for early access :)
