OpenAI open-sourced Codex Security today. It finds vulnerabilities, validates them, and patches the code. Good tool. Does one job well.
But finding and patching a bug isn't the end of the story. Who's it assigned to. Is it tracked. Does anyone follow up.
That part still lives in a different tool, a different login, re-typed into a ticket by hand.
On CREAO the same workflow can stay in one chat. The agent that finds the issue routes it straight to the agent that owns the ticket, the assignment, the follow-up. No export, no re-typing.
What's the one job you'd want an agent to own, start to finish?