TurfScope NYC
Heat pump turf map for NY Clean Heat door-to-door sales. It scores census tracts in Queens, Nassau and Suffolk (then the rest of the turf area) by how many 1–4 family homes there are, how old they are, what they heat with, owner-occupancy and incentive eligibility. Every tract is tagged Con Edison or PSEG Long Island. Queens also gets address-level walk lists.
Quick start
npm test # offline check, no network needed
npm run pipeline # pull live data + score + walk lists (a few minutes)
npm run dev # open http://localhost:3000
Requires Node 18.17+. No npm install needed — there are zero dependencies.
The pipeline needs a free Census API key (https://api.census.gov/data/key_signup.html). Put it in .env as CENSUS_API_KEY for local runs, and add it as a repository secret named CENSUS_API_KEY (Settings → Secrets and variables → Actions) for the GitHub refresh.
Team tracking (4-man team)
Reps sign in on their iPhones, claim turf (the map shows who has what), knock houses (No answer / Come back / Not interested / Interested / Booked), and leave notes and map pins for each other. Everything saves on the phone first (works with no signal), syncs to Supabase, and the whole team sees it. Setup: supabase/README.md. The team (Matt, Gio, Kai, Ash) lives in config/team.json.
Deploy
Push to GitHub, import the repo in Vercel, and deploy. vercel.json serves m3-map/ as-is.
To refresh the data without running anything locally: GitHub → Actions → Refresh data → Run workflow. It runs the tests and the live pipeline, then commits m3-map/data/, and Vercel redeploys. The office CSV (walklists_all.csv) is attached to the run as a download. You can still run npm run pipeline locally and commit the result instead.
Docs
SPEC.md — what it does and whyCLAUDE.md — how to work on it (read this first if you're an agent)config/ — every tunable number, endpoint and the turf boundary
