Quick UX feedback on agent orchestration and the sidebar IA, sharing as a user.
Three things that consistently trip me up:
1. The sidebar treats Zaps, Tables, Forms, Chatbots, Canvases, and Agents as peers, but they are different layers of the system. Primitives, surfaces, and orchestration are flattened into one list, which makes the mental model hard to form.
2. The flow starts with tools and integrations. As a user I would expect to start with a project or task, and have the relevant primitives and integrations surface in context after that.
3. The line between an Agent and an automated task (Zap, Chatbot) is not clear from the UI copy. Agent orchestration is a new space, so users are leaning hard on labels and empty states to learn the model, and right now those are not doing enough work.
Would love to see the sidebar grouped by intent or lifecycle (build, orchestrate, deliver, observe) and the agent vs automation distinction made explicit in copy. Happy to walk through specifics.
