Switch the assistant model from qwen/qwen3-32b to openai/gpt-oss-120b. Cap every per-role context list, hard-clamp the assembled system prompt to ~4000 tokens at a line boundary, and bound re-sent chat history to 12 messages so the model context window can't be exceeded regardless of mock-data volume.
- Fix tab overflow: scrollable tab bar, shortened labels (Open/Risk/Done)
- Cap DROP_POOL to 3 fires, no repeat flooding
- Resizable drag handles on desktop (xl 1280px breakpoint)
- Collapsible panels on mobile/tablet — Map + AI collapsed by default on mobile
- iPad Air + iPad Pro 11" treated as tablet via xl breakpoint
- Fixed lead list heights: 22rem mobile → 72rem xl (8 cards visible)
- PageTransition h-full fix to maintain height chain
- Add Estimate_Section_Screens_and_Workflows/ to .gitignore (local reference images only)
- Rewrite README to reflect revamp branch: React 19, ProCanvas dual-mode design, EstimateBuilder redesign, Lead Creation and LynkDispatch AI upcoming features
- Animations: Added global page transitions and smooth sidebar toggle
- Loaders: Implemented branded Loader component and page loading states
- Fixes: Resolved hook order violations in Maps/Dashboard; fixed Layout import
- Docs: Updated README with correct demo credentials and feature list
Satyam Rastogi