Wardrive Go · Live Stats

A graphical, auto-refreshed dashboard of aggregate, anonymous usage stats for the Wardrive Go app, built from the Google Analytics Data API (GA4).

• A scheduled GitHub Action (.github/workflows/stats.yml) pulls GA4 every ~30 min and commits docs/data/stats.json.
• The static dashboard in docs/ (served by GitHub Pages) renders that JSON with Chart.js.
• The Wardrive Go app's hidden stats page loads this site in a WebView.

No secrets live in this repo. The service-account key is injected at runtime from a GitHub Actions secret and is never printed or committed. Only aggregate dimensions are queried (event counts, active users, app version, country, device model, OS version) — no user- or device-level data.

Live site: https://rocketgod-git.github.io/wardrive-stats/ (after Pages is enabled + the secret is set).

Everything below is done once, in the Google Cloud / Analytics consoles and this repo's settings. The code is already wired to consume it.

1. In Google Cloud Console → the project that backs this GA4 property → IAM & Admin → Service Accounts → Create service account (e.g. ga4-stats-reader). No project roles needed.
2. On that service account → Keys → Add key → JSON → download the JSON file.
3. Enable the Google Analytics Data API for the project (APIs & Services → Library → "Google Analytics Data API" → Enable).
4. In Google Analytics (not Cloud) → Admin → Property access management → + → add the service account's email (the client_email from the JSON) with the Viewer role.

GA Admin → Property settings → the numeric Property ID (e.g. 123456789). It's the number, not the G-XXXX measurement id.

This repo → Settings → Secrets and variables → Actions → New repository secret:

• GA4_SA — paste the entire contents of the service-account JSON file.
• GA4_PROPERTY_ID — the numeric property id.

Settings → Pages → Build and deployment → Source: Deploy from a branch → Branch: main / /docs.

Actions → "Fetch GA4 stats" → Run workflow. It will fetch, commit stats.json, and Pages will publish. (Before the secret exists the run is a harmless no-op and the sample data stays on the page.)

cd docs && python -m http.server 8000   # then open http://localhost:8000

The page renders whatever is in docs/data/stats.json (the committed sample until the job runs for real).

scripts/fetch_stats.py queries only standard GA4 dimensions. To chart custom event parameters (e.g. notable_spotted.category, upload.target), register them as custom dimensions in GA Admin → Custom definitions, then add a report for them in the script (the code already isolates each report so a missing dimension can't break the file).