DRIP is a collaborative workspace platform that uses AI agents to help teams coordinate their work through a shared real-time state. The objective of this project is to extend the current DRIP prototype by introducing a structured and actionable representation of the shared state. At the moment, the system relies on a text-based snapshot, which limits its ability to support interaction and rich visualization. The goal is therefore to design and implement a new state layer that interprets user inputs as intents, translates them into structured updates, and reflects them in a dynamic interface. This structured state should directly represent the current state of the work, making it explicit, interactive, and usable as the primary surface for coordination.