Vacation Rental Turnover Coordinator

AI Prompts to Build This

Copy these build prompts into Claude, Cursor, or your AI coding tool. Create a free account to unlock the full research behind them.

1. 1. Project Setup

   Create a Next.js 14 (App Router, TypeScript, Tailwind) project called "TurnoverBoard." Set up Supabase Postgres with these tables: hosts (id, email, phone, timezone), properties (id, host_id, name, ical_url_airbnb, ical_url_vrbo, ical_url_booking, default_cleaner_id, tight_turnover_hours INT default 6), cleaners (id, host_id, name, phone, active BOOLEAN), reservations (id, property_id, source TEXT, external_id TEXT, checkin TIMESTAMPTZ, checkout TIMESTAMPTZ, guest_name TEXT), turnovers (id, property_id, prev_reservation_id, next_reservation_id, gap_hours FLOAT, is_flagged BOOLEAN, assigned_cleaner_id, status TEXT default 'pending'), checklist_items (id, property_id, label, requires_photo BOOLEAN, sort_order INT), checklist_completions (id, turnover_id, checklist_item_id, photo_url, completed_at). Enable row-level security scoped to host_id. Install the Twilio Node SDK, node-ical, and the Vercel AI SDK.

2. 2. Calendar Sync + Turnover Risk Engine

   Build a Vercel Cron job that runs every 15 minutes and, for each active property, fetches its Airbnb, VRBO, and Booking.com iCal URLs using node-ical. Parse each feed into events with UID, start date, and end date. Upsert into the reservations table, matching on (property_id, source, external_id) so re-syncs update rather than duplicate. After syncing a property's reservations, recompute its turnovers: for every checkout, find the next reservation's checkin at the same property, compute gap_hours between them, and upsert a turnover row. If gap_hours is less than the property's tight_turnover_hours, set is_flagged = true and enqueue an SMS via Twilio to the host and the assigned cleaner: "Tight turnover at {property_name}: only {gap_hours} hours between checkout and next check-in on {date}. Confirm you can make it." Also flag (separately) any cleaner who has two is_flagged turnovers at different properties overlapping in time, since they cannot physically complete both.

3. 3. Cleaner Checklist + Photo Proof Flow

   Build a no-login mobile web page at /turnover/[turnoverId] that a cleaner opens from an SMS link (use a signed, expiring token in the URL instead of auth). Show the property name, checkout/checkin times, and the property's checklist_items in order. For each item where requires_photo is true, show a camera-capture input; upload the photo to Supabase Storage and write a checklist_completions row with the photo_url and completed_at timestamp. Show a progress bar (e.g., "6 of 9 complete"). When all required items are done, mark the turnover status as 'complete', notify the host via SMS with a summary link, and show the cleaner a confirmation screen. Handle the case where a cleaner opens the link after the checkin time has already passed by showing a red "guest may already be arriving" banner at the top of the page.