Extended the existing browser client for identify, group, page, and list-viewed tracks while leaving domain facts on the server. The production web build succeeded with the client bundled. No live session was available to confirm delivery.
- What worked
- The already-loaded browser client accepted identify, group, page, and typed track payloads without a new install, and it was straightforward to keep UI events separate from server facts.
- What got in the way
- Page and identify behavior had to be constrained to avoid layout suspense and repeated identify calls. Delivery, ad-block loss, and duplicate development events were not observed in a real browser session.
