Skip to content

About

Examples for the Gridics Zoning API: parcel zoning, property search, development capacity, site selection, due diligence, appraisal, MCP and AI integrations.

Topics

Resources

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

Gridics Zoning API Cookbook

Build zoning-aware real estate applications, property workflows, and AI agents with the Gridics Public API.

Use Gridics to resolve properties, retrieve parcel-specific zoning, search and screen parcels, evaluate available development-capacity fields, enrich portfolios, support appraisal and due diligence, add address/APN search to applications, and connect property data to AI clients through Model Context Protocol (MCP).

Create a Developer account · Developer cookbook hub · Gridics Zoning Data API · Browse the API contract · Choose a use case

Verification status: published as a preview cookbook. All seven language implementations pass credential-free fixture tests. End-to-end production acceptance and ordinary developer onboarding are still incomplete; examples are not guaranteed to work for every plan or geography. Location Search remains preview until suggest/retrieve/forward and the seven-language production matrix pass. Fixture success is not production verification. See the recipe verification status for the latest local audit.

What can you build?

Gridics is designed for applications that need structured property and zoning intelligence rather than a generic property record alone.

  • Property zoning lookup — resolve an address, APN, parcel ID, or coordinates and retrieve zoning for the canonical parcel.
  • Parcel and land-use search — find properties with simple, advanced, or polygon-based filters.
  • Development opportunity screening — discover released fields first, then use parcel and development-capacity fields where your plan and geography expose them.
  • Site selection — create explainable property shortlists using supported parcel and zoning filters.
  • Portfolio enrichment — add parcel, zoning, and property facts to existing property lists.
  • Appraisal and due diligence — produce factual subject-property zoning summaries with provenance and explicit unknowns.
  • Location Search — add address/APN suggest, retrieve, and forward search to an application.
  • AI and MCP — expose bounded Gridics tools to AI applications and agent workflows without putting API keys in prompts.

See a working browser example

The React Zoning Explorer resolves a selected address or APN and shows the parcel beside its zoning information. Run its synthetic demo without an account or API key, then connect your own server adapter when you are ready to use entitled live data.

Synthetic Zoning Explorer preview showing property identity beside zoning facts.

Preview uses synthetic data. The zoning panel is explicitly labelled and does not represent a live zoning determination.

Build with Gridics

I want to... Start here
Get zoning for an address or APN Address to zoning
Resolve one free-form address or APN Location Search forward
Add interactive autocomplete Location Search suggest → retrieve
Inspect abutting parcel zoning and centroid directions Abutting parcels example
Look up an exact known parcel Property lookup → Parcel zoning
Search parcels with common filters Simple search
Search with verified typed filters Field discovery → Advanced search
Build a property zoning lookup Property zoning guide
Add address autocomplete to an app React Location Search or plain browser example
Show zoning and development information in a product UI Zoning Explorer
Find development or redevelopment candidates Advanced search → Polygon search → Redevelopment screen
Build a real-estate site-selection workflow Retail/site shortlist and site-selection guide
Enrich a portfolio Portfolio enrichment
Produce a property/zoning factsheet Property factsheet
Detect changes in customer-managed property snapshots Snapshot diff
Build a property due-diligence workflow Due diligence
Add zoning to appraisal software Appraiser zoning lookup
Connect an AI client with MCP Gridics MCP integration guide
Build an AI zoning assistant AI zoning assistant guide

The complete recipe index includes difficulty, endpoints, required capabilities, and implementation status.

Get an API key

1. Create a Developer account

Open the Gridics Developer Portal, create or join an Organization, choose the geography and plan appropriate for your use case, and create an API key under Workspace → API Keys.

Your Organization owns its API keys, subscriptions, geography access, and usage. The Developer Portal is authoritative for current plans, quotas, pricing, field availability, and product access.

2. Configure the cookbook

git clone https://github.com/gridics/zoning-api-cookbook.git
cd zoning-api-cookbook
cp .env.example .env

Edit .env to set GRIDICS_API_KEY and then load the variables into your shell before running a live recipe:

set -a
source .env
set +a

Never commit a secret key, embed it in browser code, place it in a URL, or paste it into an AI prompt.

3. Verify your key

curl --fail-with-body \
  --header "x-api-key: ${GRIDICS_API_KEY}" \
  "${GRIDICS_API_BASE_URL:-https://api.gridics.com}/v2/me"

Or run the first Python recipe:

cd recipes/01-hello-world/python
python3 main.py

4. Discover your available geography

Run recipe 02 to distinguish Gridics coverage from the counties your Organization is entitled to query.

5. Solve a real problem

For a one-request zoning result from an address or APN, use recipe 19. Interactive UI uses suggest → retrieve so a person chooses a candidate; one-off resolution without zoning uses forward. The older multi-step query remains in recipe 04. Ambiguous results require a choice before proceeding. For development/site-selection workflows, start with recipe 05 or the verified filters in recipe 08.

Why Gridics for zoning-aware applications?

Many real-estate APIs stop at parcel and assessor facts. Gridics combines property resolution with parcel-specific zoning data and exposes richer development-related fields where they are released for the caller's plan and geography.

The cookbook demonstrates how to work with those capabilities safely:

  • discover fields and capabilities before relying on them;
  • preserve unknown, unavailable, unsupported, and not-entitled states;
  • keep canonical Market, Place, parcel, and APN identifiers intact;
  • retain request/provenance information for professional workflows;
  • bound pagination, retries, usage, and result counts;
  • keep server credentials out of browsers and AI tool arguments.

See the Zoning API guide, development capacity guide, and parcel zoning data guide.

Recipes

Stage Recipes What you learn
First request 01, 02 Verify a key, understand coverage vs entitlement, and select canonical Market/Place IDs.
Property + zoning 03, 04, 18 Resolve a property and retrieve parcel-specific zoning.
Search 05–09 Search, paginate, handle errors, discover fields, and search polygons.
Business workflows 10–15, 17 Portfolio, redevelopment, site selection, factsheets, change detection, due diligence, and appraisal.
AI integrations 16 Build a bounded MCP-compatible tool surface over Gridics data.

Every recipe has equivalent implementations in Python, Node.js JavaScript, TypeScript, PHP, C#/.NET, Java, and Go. Basic HTTP operations also include cURL examples.

Featured example: Zoning Explorer

examples/react-zoning-explorer is the showcase UI for composing a property-search experience with a zoning/development panel. It runs with synthetic fixtures by default and deliberately requires a server-side zoning adapter for live zoning calls so secret API keys never enter browser JavaScript.

Gridics Zoning Explorer preview

MCP and AI agents

Gridics Public API MCP uses the canonical /mcp connection exposed by the deployment and the Developer Portal's MCP Connections experience. Use the MCP integration guide for architecture, client connection guidance, and safe tool design.

For AI applications, prefer bounded typed tools over pasting zoning documents or credentials into prompts. Resolve a canonical property first, call Gridics tools, preserve references and unknowns, and let the model explain returned facts rather than invent missing zoning values.

Search and technical guides

These guides are intentionally useful standalone entry points for developers evaluating zoning and property-data solutions:

Language support

Language Runtime
Python 3.11+
Node.js JavaScript 20+
TypeScript TypeScript 5 / Node.js 20+
PHP 8.2+
C#/.NET .NET 8+
Java 21+
Go 1.22+

The language runners share one reviewed recipe manifest but remain native implementations. CI checks all recipe entry points, request parity, secret redaction, fixture behavior, the browser examples, and the React examples.

Common configuration

Variable Required Meaning
GRIDICS_API_KEY Yes for live server calls Organization secret API key. Never expose it in a browser or AI prompt.
GRIDICS_API_BASE_URL No Defaults to https://api.gridics.com.
GRIDICS_MARKET_ID Market-scoped recipes Canonical ID returned by /v2/markets.
GRIDICS_PLACE_ID County/Place-scoped recipes Canonical entitled Place ID.
GRIDICS_PARCEL_ID Zoning recipes Canonical parcel/group identifier.
GRIDICS_DRY_RUN No Set to 1 to inspect redacted request plans.
GRIDICS_FIXTURE_MODE Tests only Synthetic local fixture mode.

The values in .env.example are synthetic and will not work against production.

Test without an API key

python3 tests/validate_repository.py
python3 tests/run_fixture_smoke.py --language python
python3 tests/run_fixture_smoke.py --language nodejs
python3 tests/run_fixture_smoke.py --language go

Fixture tests make no live or billable requests. Protected live verification is a separate bounded workflow; fixture success is never reported as production verification.

Data and professional boundaries

Gridics returns factual data according to the caller's active capabilities and geography. A cookbook example is not a legal zoning determination, formal appraisal, title opinion, credit decision, availability claim, or assurance of permit approval. Missing values remain missing; null is not zero.

See request conventions, compatibility notes, security guidance, and support.

Contributing

Contributions are welcome when they use only public Gridics interfaces, public dependencies, and synthetic or explicitly approved examples. Read CONTRIBUTING.md and AGENTS.md first.

Licensed under the Apache License 2.0.

About

Examples for the Gridics Zoning API: parcel zoning, property search, development capacity, site selection, due diligence, appraisal, MCP and AI integrations.

Topics

Resources

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Used by

Contributors

Languages