clean-room
Remove specified clutter from a room photograph with an AI-native contextual edit while preserving architecture, fixed furniture, perspective, materials, lighting, and the room's identity. Use for rea
SKILL.mdskills/clean-room/SKILL.md
GitHubClean Room
Catalog
- Display name: Clean Room
- Categories: Utility
- Status: Ready
- Route label: GPT Image 2 consistency-sensitive edit
- Tone: green
- Short description: Remove room clutter without redesigning the space.
- Similar skills: remove-object, restyle-room
What this skill does
- Removes a declared clutter set as one coherent cleanup result.
- Reconstructs floors, walls, textiles, and surfaces while preserving architecture and fixed furnishings.
How to use
Provide room photo and list removable clutter versus protected furniture and decor.
Clean a bedroom
Remove clothes, cables, cups, and loose papers; keep all furniture, plants, art, bedding, and room layout.
Workflow
- Inventory removable clutter and protected structure, furniture, decor, textiles, windows, doors, and fixtures.
- Run
gpt-imagev1.0 withquality: "medium",imageSize: "2K", the supplied references, the requested or source-preserving ratio,batchCount: 1, and one result. - Request one spatially coherent cleanup; forbid redesign, new furniture, staging, material changes, or relighting.
- Inspect perspective lines, floor/wall continuation, reflections, shadows, repeated textures, furniture count, and room proportions.
- Reject missing protected items, added decor, warped architecture, floating furniture, or over-empty staged results.
User-facing output
- Media type: One cleaned room image
- Default quantity: 1
- Content per image: Same room with specified clutter removed
- Default layout: Source composition
- Model policy: GPT Image 2 Medium/2K for consistency-sensitive final image editing
- Downstream use: Real-estate, hospitality, and interior listings