Cronny

Schedule the web checks that matter to you.

Cronny is a self-hosted scheduler for monitoring websites and receiving useful results on a schedule. Create jobs in a web UI, keep their history in SQLite, and get notified when something changes. It includes ready-to-use monitors for Austrian and Bosnian real estate, public appointments, reservations, and news feeds.

Highlights

• Run on your schedule: standard cron expressions, enabled/disabled jobs, next-run times, and run history.
• Keep the signal: preserve source search URLs and refine real-estate results by price, size, rooms, radius, or a map-drawn polygon.
• Get notified: Slack and WhatsApp notifications can be sent after successful or failed runs; successful notifications can be limited to new results.
• Use the built-in sources: monitor real estate listings, ID Austria appointments, Vienna grill reservations, generic Willhaben listings, and AI-processed RSS news.
• Own the data: jobs, results, runs, clients, and schedules are stored locally in SQLite.
• Share selected results: issue API keys to clients and grant access only to specific jobs.

Supported monitors

Category	Sources
Real estate	DerStandard Immobilien, Willhaben Immobilien, ImmoScout24 Austria, OLX.ba, and a multi-source Immo monitor
Availability	ID Austria appointments and Vienna public grill-area reservations
News	Klix.ba and DerStandard RSS feeds with optional OpenRouter AI processing
General	Generic Willhaben searches

Quick start with Docker

The Docker setup runs Cronny and FlareSolverr, which helps with sources that require browser-like challenge handling. Docker Compose persists Cronny's SQLite data in a named volume.

1. Create packages/server/.env:

   API_SECRET=replace-with-a-long-random-secret
   API_PASSWORD=replace-with-a-strong-password
   BASE_URL=http://localhost:3030
   
   # Required when using location-based real-estate filters
   GEOCODE_API_KEY=
   
   # Required for AI news monitors
   OPENROUTER_API_KEY=
   
   # Required only when using WhatsApp notifications
   WHATSAPP_API_KEY=

2. Start the services:

   docker compose up --build -d

3. Open http://localhost:3030, then sign in with API_PASSWORD.

Check that the service is ready:

curl http://localhost:3030/api/public/health
# {"status":"ok"}

To stop the stack, run docker compose down. Add -v only if you also want to delete the stored database.

Local development

Prerequisites

• Node.js 22.14.0
• pnpm 9.15.4 (Corepack is recommended)
• Docker (optional, for FlareSolverr)

corepack enable
pnpm install --frozen-lockfile

Create packages/server/.env with the values from the Docker quick start, plus the local server settings:

NODE_ENV=development
PORT=3030
API_SECRET=replace-with-a-long-random-secret
API_PASSWORD=replace-with-a-strong-password
BASE_URL=http://localhost:5174

Start both applications:

pnpm dev:all

• Web app: http://localhost:5174
• API: http://localhost:3030

For sources that need FlareSolverr, start it separately:

docker compose up flaresolverr

The server creates its development database at packages/server/.data/db.sqlite. Its development startup also generates and applies Drizzle schema changes.

Configuration

Variable	Required	Purpose
NODE_ENV	Local development	Set to development locally; Docker sets production automatically.
PORT	Yes	HTTP port for the server.
API_SECRET	Yes	Secret used to sign web-session tokens. Use a long, random value.
API_PASSWORD	Yes	Password for the web interface.
BASE_URL	For notifications	Public URL used in notification links.
GEOCODE_API_KEY	For geocoded sources and geo filters	API key for address geocoding.
OPENROUTER_API_KEY	For AI news monitors	OpenRouter API key for RSS processing.
WHATSAPP_API_KEY	For WhatsApp notifications	CallMeBot API key.
Job notification phone	For WhatsApp notifications	Configure the recipient phone number in the job's notification settings.
FLARESOLVERR_URL	Optional	FlareSolverr endpoint; defaults to http://localhost:8191.
DATABASE_PATH	Optional	Override the default SQLite database path.

Never commit .env files or credentials.

Commands

Run these from the repository root:

Command	Description
pnpm dev	Start the server in watch mode.
pnpm dev:all	Start the client and server together.
pnpm build	Build shared types, server, and client.
pnpm lint	Lint the client and server.
pnpm test	Run server tests.
pnpm start:prod	Run migrations and start the production server.

Architecture

packages/
├── client/  React, Vite, Mantine, TanStack Query/Router
├── server/  Hono, Drizzle, SQLite, Playwright, node-cron
└── types/   Shared TypeScript contracts

Each monitor is a strategy in packages/server/src/strategies/. Strategies share a runner contract; real-estate strategies additionally share filtering and geographic matching logic. The server schedules persisted jobs with cron, records each run and its results, and serves the client and API.

API access

Cronny can expose results to external clients with API keys. Create a client and assign its permitted jobs in the UI, then request:

curl -H "X-API-Key: YOUR_CLIENT_KEY" \
  http://localhost:3030/api/public/results

The endpoint returns results only from jobs assigned to that client.

Contributing

Issues and pull requests are welcome. Please use pnpm, keep changes focused, and run these checks before opening a pull request:

pnpm lint
pnpm build
pnpm test

License

MIT