Gwinnett Index MCP server

Gwinnett Index

A machine-readable index of zoning, land-use, and development records for Gwinnett County, Georgia and its 17 municipalities. Built for AI agents first, humans second.

Why this exists

A "Duluth, GA" mailing address is usually not in the City of Duluth. It's in unincorporated Gwinnett — a different code, a different board, a different permit portal. Nothing on the internet resolves an address to its governing jurisdiction before answering a zoning question, so every AI assistant answering "what's the setback in Duluth GA" today is guessing.

Unincorporated Gwinnett is 67.3% of the county's land area. Getting this wrong is the default, not the edge case.

What's here

Setup — one time

The repo already has a starter README, so clone first and copy in — that avoids a push conflict:

git clone https://github.com/willmobhill-arch/gwinnett-index.git
tar -xzf gwinnett-index-repo.tar.gz -C gwinnett-index      # overwrites the stub README
cd gwinnett-index
git add -A && git commit -m "Gwinnett Index: pipeline, schema, and extracted corpora"
git push

Then, with the repo public, load the corpora. Postgres fetches them itself — you don't run anything locally, and no credential is involved:

-- set to your repo's raw base
-- https://raw.githubusercontent.com/willmobhill-arch/gwinnett-index/main/data/

SELECT load_udc_sections_from_url(
  'https://raw.githubusercontent.com/willmobhill-arch/gwinnett-index/main/data/udc_sections.jsonl');

SELECT * FROM load_duluth_cases_from_url(
  'https://raw.githubusercontent.com/willmobhill-arch/gwinnett-index/main/data/duluth_cases.jsonl');

SELECT load_meeting_docs_from_url(
  'https://raw.githubusercontent.com/willmobhill-arch/gwinnett-index/main/data/duluth_meeting_docs.jsonl');

Expected: 860 sections, 228 case rows (109 distinct), 356 meeting documents.

Architecture

Ingestion runs inside Postgres via the http extension — it fetches the county's ArcGIS REST services, Census TIGERweb, and this repo's own raw URLs directly. No external worker, no credentials in transit, and the whole 11,739-case county load moves zero bytes through a client.

ingest/
  duluth/          UDC PDF -> 860 sections + 18 tables (extract_udc, extract_tables)
  duluth_agendas/  crawl -> OCR -> parse case items from agendas/minutes/packets
db/migrations/     the schema, md5-verified against the live Supabase project
scripts/           export_snapshot.py (the only thing that talks to Postgres) + loaders
data/              extracted corpora, fetched by the SQL loaders above
site/              Astro static site + full agent surface (.md twins, llms.txt, DCAT)
worker/            Cloudflare Worker: REST API + MCP server, five tools, no auth
docs/plans/        design doc, build plan, status

Building the site

DATABASE_URL=postgresql://... python3 scripts/export_snapshot.py   # writes data/snapshot/
cd site && npm ci && SITE_URL=https://your-domain npm run ci        # build + 15 gates

Without a snapshot the site builds from site/fixtures/ — a small committed sample with identical shapes — and says so on every page. npm run ci fails the build if any route lacks a .md twin, any page presents records before naming its governing jurisdiction, or the bulk export contains geometry.

The API and MCP server

Shortened. The full README is on GitHub.