Kumospace?

Build a focused web prototype called a 'spatial virtual office' inspired by the Kumospace workflow: remote team members navigate a 2D top-down office map with an avatar; their avatar's position determines who they hear via simulated spatial audio and who sees their screen shares. Core workflow: a user joins a space, sees rooms and colleagues on a map, moves by clicking/arrow keys, and audio fades in/out based on distance; entering an office or conference room creates a private audio zone; a door toggle makes the room private. Requirements: real-time presence sync (WebSocket or WebRTC data channels), proximity-based audio panning/gain using Web Audio API (no external audio server), simultaneous screen sharing via getDisplayMedia with per-user audio zones, chat by room/floor, clickable link tablets and a simple persistent whiteboard canvas, admin settings for password protection and email-allowlist, and basic presence analytics recorded server-side. Implementation approach: React + Node.js/Socket.io for sync, a simple tile-based map renderer (Canvas or Phaser) for movement and collision, and WebRTC mesh for audio/video streams. Explicitly do not copy the Kumospace brand, logo, UI layout, or proprietary visual assets; design an original map and interface. Out of scope: production scalability, recording, SSO, mobile apps, virtual games, furniture customization, white-labeling, and advanced security. Deliver a README describing setup, architecture choices, known limitations (e.g., mesh scaling, Echo cancellation, browser permission handling), and a disclaimer that this is an original replacement inspired only by the general virtual-office workflow, not affiliated with Kumospace.