Pi-Schedule

A Flask + SQLite calendar, task management, and job-hunting app designed to run permanently on a Raspberry Pi. Manages calendar events (with Outlook sync), prioritised tasks with scheduled time blocks, and an automated job scraper that feeds validated listings into a separate job-applier pipeline.

Features

• Calendar — Month, Week, and Day views; manual events plus Outlook sync via Microsoft Graph API
• Tasks — Create tasks with priority (High/Medium/Low), due dates, estimated duration, recurrence rules, and sub-tasks
• Scheduled Blocks — Assign calendar time blocks to tasks; track scheduled vs. completed minutes
• Job Searches — Define keyword searches (DuckDuckGo) and direct job board URLs to spider
• Job Scraper — Runs on a schedule; verifies listings are active (heuristic keyword scoring + deadline extraction), deduplicates, and stores results
• Job Applier Integration — Each verified listing is POSTed to a configurable webhook (JOB_APPLIER_URL) for automated downstream processing
• Local-first — Everything runs on your network; SQLite database, no cloud dependency beyond optional Outlook sync

Setup (Raspberry Pi)

git clone <repo-url> ~/pi-schedule
cd ~/pi-schedule
cp .env.example .env   # edit with your settings (see below)
bash setup_pi.sh

setup_pi.sh creates a Python virtual environment, installs dependencies, and registers a systemd service (pi-schedule.service) that starts on boot.

Running

Action	Command
Start service	sudo systemctl start pi-schedule
Stop service	sudo systemctl stop pi-schedule
Check status	sudo systemctl status pi-schedule
View logs	journalctl -u pi-schedule -f

The app listens on port 5000 and is accessible at http://<pi-ip>:5000.

Local development

python -m venv venv
source venv/bin/activate      # Windows: venv\Scripts\activate
pip install -r requirements.txt
cp .env.example .env
python run.py

Environment Variables (.env)

Variable	Required	Description
SECRET_KEY	Yes	Flask session secret
AZURE_CLIENT_ID	Outlook sync	Azure app client ID
AZURE_CLIENT_SECRET	Outlook sync	Azure app client secret
AZURE_TENANT_ID	Outlook sync	Azure directory tenant ID
AZURE_REDIRECT_URI	Outlook sync	e.g. http://<pi-ip>:5000/auth/callback
JOB_APPLIER_URL	Job pipeline	Webhook URL for the job-applier service

Outlook Sync Setup

1. Go to Azure Portal → Microsoft Entra ID → App registrations → New registration
   • Redirect URI: http://<your-pi-ip>:5000/auth/callback
2. Copy Application (client) ID → AZURE_CLIENT_ID and Directory (tenant) ID → AZURE_TENANT_ID
3. Under Certificates & secrets, create a new client secret → AZURE_CLIENT_SECRET
4. Under API permissions, add Microsoft Graph → Delegated → Calendars.Read and grant admin consent

Job Scraper

The scraper runs automatically via APScheduler. It processes two source types:

• Keyword searches — queries DuckDuckGo and verifies each result URL is an active job posting
• Job boards — spiders a direct board URL for job links, then cross-references against active keyword searches

Verified listings are saved to the database and (if JOB_APPLIER_URL is set) forwarded to the job-applier webhook with title, company, description, location, salary, and deadline.

Deploy Scripts (local only, gitignored)

Three helper scripts exist locally for deploying to the Pi over SSH — they are not committed to the repo:

Script	Purpose
check_pi.py	Verify Pi is reachable and service is running
deploy_to_pi.py	Sync code to the Pi and restart the service
finish_deploy.py	Post-deploy checks

Set the PI_PASS environment variable before running them:

PI_PASS=yourpassword python deploy_to_pi.py

Project Structure

├── app/
│   ├── __init__.py        # Flask app factory + APScheduler
│   ├── models.py          # Task, Event, ScheduledBlock, JobSearch, JobBoard, JobListing
│   ├── routes_api.py      # REST API endpoints
│   ├── routes_auth.py     # OAuth2 / Outlook auth routes
│   └── sync.py            # Microsoft Graph sync logic
├── scripts/
│   └── scraper.py         # Job scraper (DuckDuckGo + board spidering)
├── static/
│   ├── css/app.css
│   └── js/                # Vanilla JS SPA (calendar, tasks, jobs views)
├── templates/
│   └── index.html         # SPA shell
├── run.py                 # Entry point
├── setup_pi.sh            # Pi systemd service installer
└── requirements.txt

Tech Stack

• Backend: Python, Flask, SQLAlchemy, SQLite, Flask-APScheduler
• Frontend: Vanilla JavaScript, Tailwind CSS (CDN), Lucide Icons (CDN)
• Scraping: Scrapling (adaptive HTTP + headless browser), BeautifulSoup, DuckDuckGo Search
• Outlook integration: MSAL, Microsoft Graph API