You’re Absolutely Right (and another Lies My AI Told Me): Engineering Context for Agentic AI
My AI coding agent agrees with me a lot. It agrees erstwhile I’m right. It agrees erstwhile I’m wrong. It agrees while deciding on its own to remove validation logic, rewrite business rules, and confidently explain why this is an improvement.
This is what coding with AI agents looks like erstwhile context lives in the developer’s head alternatively than in the system.
This talk looks at how developers can make agents little agreeable and more effective by treating shared context as an engineering problem. We’ll show how prompts, rules, and another forms of strategy context and shared cognition can be packaged into explicit, reusable units alternatively of surviving in ad hoc conversations. More importantly, we’ll look at how that context can be shared and reused consistently across tools and teams, utilizing acquainted distribution ideas borrowed from the way we already share libraries via registries and repositories.
On top of that foundation, we’ll cover erstwhile deterministic scripts are the better choice and how to embed them into agent workflows, how to measure prompt quality utilizing concrete signals alternatively of “trust me, bro,” and how guardrails like tests and structure aid agents neglect loudly alternatively of silently drifting.
If your AI keeps agreeing with you even erstwhile it’s doing the incorrect thing, you’re absolutely right: this talk is for you.
