Getting started

Requirements

Use Python 3.13 or newer and Git. Clone the repository, create a virtual environment, and install the runtime dependencies.

=== "Windows PowerShell"

```powershell
py -3.13 -m venv .venv
.\.venv\Scripts\Activate.ps1
python -m pip install -r requirements.txt
```

=== "macOS / Linux"

```bash
python3.13 -m venv .venv
source .venv/bin/activate
python -m pip install -r requirements.txt
```

First analysis

python main.py reads data/listings.json. A feed is a JSON array of listing objects with an ID, title, description, asking price, URL, location, and optional distance. The bundled fixture is safe for trying the deterministic pipeline. See Analysis before using your own data.

Web application

uvicorn web.app:app --reload

Open http://127.0.0.1:8000. Set FLIPPER_INVENTORY_DB to select another local SQLite database. The process initializes and migrates that database. Use a copy when experimenting.

Open /deals to run an explicit, read-only eBay discovery search when application credentials and production access are configured. The dashboard and other local pages do not trigger discovery.

Optional integrations

Environment variables can enable OpenAI enrichment, market-data lookup, Discord alerts, and eBay seller APIs. None is required for local deterministic parsing, heuristic pricing, inventory, or reports. Consult Configuration and never commit .env or tokens.