BuyDown Indy

A public, mobile-first map of greater-Indianapolis listings where the seller pays concessions. For each listing, buyers see what that money does as a temporary buydown (1-0, 2-1, 3-2-1), a permanent buydown, or a closing cost credit, compared with a price cut of the same amount.

Stack: Next.js 15 (App Router) · TypeScript · Tailwind v4 · Leaflet · Supabase (Postgres, Auth, Storage) · Zod · Vercel.

Run it

npm install
npm run dev        # http://localhost:3000
npm test           # payment-engine + PMMS parser tests

With no Supabase env vars the app runs in demo mode: the map and listing pages use the 12 sample listings and the 6.25% sample rate, the lender form validates but doesn't store, and agent screens explain that setup is needed.

Set up Supabase

1. Create a project at supabase.com.
2. SQL editor → run supabase/migrations/0001_init.sql, then supabase/seed.sql (sample listings + rate).
3. Authentication → URL Configuration: set Site URL to the production URL and add https://<your-domain>/auth/callback (and http://localhost:3000/auth/callback) to Redirect URLs.
4. Copy the URL, anon key and service-role key into .env.local (see .env.example) and into Vercel.

Remove the demo data later with delete from listings where is_sample;.

Lead emails

Every lead is saved, then emailed to LEAD_NOTIFY_TO through Resend (src/lib/lead-email.ts) with the quiz answers and consent record. Reply-To is the buyer, so replying answers them directly. With no RESEND_API_KEY the email is skipped; the lead is still saved.

Daily rates

Admins (emails in ADMIN_EMAILS) enter Rate's 30-year fixed Conventional, FHA and VA rates at /agent/rates. They're used until midnight Indianapolis time. On any day without an entry the site falls back to the latest Freddie Mac PMMS weekly average, labeled a sample rate. In local demo mode anyone on localhost can enter rates and they're saved to .data/daily-rates.json.

Deploy (Vercel)

Import the GitHub repo in Vercel, add the env vars from .env.example, deploy. vercel.json schedules /api/cron/rate every Thursday at 18:00 UTC to pull the latest Freddie Mac PMMS 30-year rate.

Where things live

Path	What
src/lib/buydown.ts	Payment engine (port of the prototype's calc()), tested in buydown.test.ts
src/content/disclosures.ts	All compliance copy (disclosures, TCPA consent, footer). Edit here only.
src/content/program-rules.ts	Loan program rules from Rate's official Buydown & IPC Calculator: concession limits, MI/MIP, FHA UFMIP, VA funding fee, 4%-of-loan closing-cost estimate
src/lib/rate-calc-parity.test.ts	Checks the engine against Rate's calculator formulas across 96 scenarios
src/components/map/	Map screen: Leaflet wrapper, bottom sheet, filters, hero
src/components/listing/	Listing page, option cards, gallery
src/content/lender-quiz.ts	"Talk to a lender" quiz questions and follow-ups (edit here)
src/components/lender/	The quiz at /talk-to-a-lender?listing=<id>[&topic=points]
src/components/agent/	Posting wizard, photo grid (dnd-kit), dashboard, login
src/app/api/	leads, listings, agent, geocode, cron/rate
supabase/	Schema, RLS policies, storage bucket, seed

Before public launch

Search the code for COMPLIANCE:. Open items:

• Rate compliance review of all copy in src/content/disclosures.ts.
• Representative example likely needs a real APR (Reg Z trigger terms).
• TCPA consent wording in the lender form.
• Advertised-rate rules: the PMMS average is labeled a sample rate.
• EHO logo: src/components/ui/EhoLogo.tsx is drawn to match HUD's mark; confirm or swap in the official file.
• A production tile provider key (NEXT_PUBLIC_MAPTILER_KEY or NEXT_PUBLIC_TILE_URL).
• Lead emails: set RESEND_API_KEY, LEAD_NOTIFY_TO and a verified LEAD_NOTIFY_FROM (see .env.example).
• Privacy policy: the quiz collects income, debts and credit range, so a privacy policy should be linked before launch.