GPT-6 Astra builds a full hospital corridor in Blender via MCP in ~45 minutes
Time-Ad-7720 · reddit · 2026-09-13
An art director (not a 3D artist) used GPT-6 Astra with Blender MCP to build a complete 35m game-ready hospital corridor entirely from text prompts — architecture, modular door variants, patient rooms, props (beds, IV stands, wheelchairs), a nurse station, lighting, and a cinematic flythrough — in about 45 minutes, with no manual modeling and no external assets.
The workflow ran in four approval-gated stages:
- Blockout: architecture only, organized collections, cinematic camera setup;
- Dressing: reusable door panels, deliberate prop placement across foreground/midground/background, environmental storytelling (loose papers, displaced ceiling tile), reference-matched lighting, preview render;
- Camera animation: an 8-second eased push-in with subtle handheld sway, geometry untouched;
- Relighting: AgX color management, darker foreground pools, warm light from patient rooms, subtle volumetric haze.
The key lesson: split the job into stages and pause for human approval between each, instead of letting the model one-shot the whole scene. The author offered to write a full setup tutorial.
More from coding & agent
- Game Studio Uses AI Agents to Play Multiplayer Games and File QA Reports — draginol · 2026-10-31
- Stardock's QA now runs agents that play the game and file reports — draginol · 2026-10-31
- Limma MCP structures unstructured data end-to-end so agents never miss a detail — FlimsyIngenuity7248 · 2026-10-31
- Agent audit metric: can you name which tests ran and files changed without reading its summary? — ronbodkin · 2026-10-31
- FlowSense MCP lets Claude Code inspect and firewall your Mac's network traffic locally — hani_q · 2026-10-31
- Dev: coding agent notifications only worth pushing when waiting on approval — petrusenko_max · 2026-10-31