Integration

MCP surface

the305Home Studio is designed to be driven from ChatGPT. These are the tools the MCP server will expose once the Supabase project is connected — the same operations the UI performs.

list_drops

{ status?, city?, from?, to? }

Return drops with their status, space, design direction and scheduled date.

get_drop

{ dropId | code }

Full brief: design recipe, selected room concept, candidates, staging notes.

create_drop

{ date, city, residenceType, space, designDirection, recipe }

Open a new slot on the bi-weekly calendar.

propose_products

{ dropId, products[] }

Push sourced candidates in from research with pricing, retailer, availability and score inputs.

score_product

{ productId, edge, cohesion, value, affiliate, availability, delivery }

Set or override the 305 Score inputs for a piece.

set_candidate_status

{ candidateId, status, tier?, note? }

Move a candidate between Proposed, Selected, Saved and Rejected.

write_storyboard

{ dropId, templateId, scenes[] }

Write the 30-second scene list for the reel.

queue_post

{ dropId, scheduledDate, caption, hashtags[], cta }

Add or update the publish queue entry for a drop.

Drop JSON import

Until the MCP server is live, research done in ChatGPT can be handed over as JSON. Paste a payload with a top-level drops array.