📋 Muster

Attendance & Presence Platform — present, and accounted for.

Location- and face-verified attendance for many organizations at once. Employees check in from a mobile app; admins manage staff, locations, shifts, and reports from a web dashboard. Multi-tenant, co-branded, and built entirely on free tiers — no credit card.

✨ Features · 🖼️ Screenshots · 🧭 Architecture · ⬇️ Setup · 🗺️ Roadmap

---

Important

Educational & Authorized Use Disclaimer
This project is developed strictly for educational purposes, legitimate organizational management, and authorized attendance tracking. The developer (@n0xnull) does not condone unauthorized deployment or privacy violations. See DISCLAIMER.md for full terms.

🇮🇩 Baca dalam Bahasa Indonesia

Muster — Platform Presensi & Kehadiran

Muster adalah platform presensi multi-tenant: satu backend dan satu aplikasi melayani banyak organisasi sekaligus. Karyawan check-in dari aplikasi mobile; admin mengelola staf, lokasi, shift, dan laporan dari dashboard web.

Verifikasi kehadiran: lokasi (geofence Haversine, radius per lokasi) dan wajah (MobileFaceNet on-device — hanya vektor yang dikirim, bukan foto). Deteksi GPS palsu (mock location) bawaan.

Fitur: Import karyawan via Excel, penjadwalan massal, ekspor CSV, super admin untuk mengelola banyak institusi, 100% gratis (Supabase + Vercel + Expo).

⚠️ Deploy hanya dalam organisasi yang Anda miliki/kelola dengan persetujuan karyawan. Lihat DISCLAIMER.md.

🧩 The Problem

Attendance across many organizations usually means a separate app per institution, easy buddy-punching (titip absen), and real cost. Field workforces — ambulance drivers, security posts, field teams — are especially hard to track because they report from many locations, not one office.

Muster is a single multi-tenant platform: one backend and one app serve many institutions. Presence is verified by location (geofence) and face, while admins manage everything from a web dashboard. To muster is to assemble personnel and take the roll — to confirm who is present and accounted for. That is exactly what this platform does, for every organization that plugs into it.

✨ Key Features

• 📍 Location check-in (no paid Maps API) — Haversine geofence; radius set per location by the admin; check-in rejected outside the radius, with a live map view.
• 🛰️ Anti-fake-GPS — detects mock-location apps and flags spoofed coordinates.
• 🕘 Shift-aware — check-in/out tied to the day's shift; on-time / late computed automatically, with a personal attendance history.
• 🧑‍💼 Face verification — on-device liveness (anti-photo) + identity matching with MobileFaceNet embeddings. Private: only vectors leave the device, not photos.
• 🏢 Multi-tenant by design — data isolated per organization via PostgreSQL Row Level Security. One database, many institutions.
• 🎨 Co-brand — the Muster brand stays primary; each institution appears as a secondary label (Muster · Institution Name) with its own small logo.
• 🗂️ Full management — employees (manual + Excel import), locations, shifts, and schedules (bulk assign to many employees across a date range).
• 📊 Reports & recap — filter by date/employee, on-time/late/mock summary, CSV export for Excel.
• 🏛️ Super admin — create and manage institutions with no redeploy.
• 🆓 100% free tier — Supabase, Vercel, Expo/EAS, OpenStreetMap. No credit card.

🖼️ Screenshots

📱 Mobile — employee app

_{Login by NIP · Home with live GPS geofence + shift status · Photo check-in / check-out · Attendance history}

🖥️ Web admin — institution dashboard

_{Dashboard · Employees · Locations & radius · Shift scheduling · Reports + CSV export · Co-brand identity}

🧭 Architecture

              MOBILE (employee)            WEB ADMIN (institution)
              React Native · Expo          Next.js · Tailwind
              check-in: GPS + face         staff, locations, shifts,
              dynamic co-brand             Excel import, reports
                     \                         /
                      \        HTTPS · JWT    /
                       \                     /
                     ┌───────────────────────────┐
                     │   BACKEND — Supabase       │
                     │   PostgreSQL + RLS         │
                     │   Auth · Storage · Edge fn │
                     │   every row → tenant_id    │
                     └───────────────────────────┘

Monorepo layout:

muster/
├── supabase/     # schema + RLS + auth functions + storage policies + seed
├── apps/
│   ├── admin/    # web dashboard (Next.js)
│   └── mobile/   # employee app (Expo / React Native)
└── assets/       # brand + screenshots

Full technical design: ../MUSTER_BLUEPRINT.md.

⬇️ Setup

1) Backend — Supabase (do this first)

Follow supabase/README.md: create a free Supabase project, run migrations/0001 → 0002 → 0003 → 0004 → seed.sql, create the Storage buckets (attendance-photos private, branding public), and turn off Authentication ▸ Email ▸ Confirm email (employee login uses virtual NIP@muster.com addresses). Copy the Project URL and anon key.

2) Web admin — Next.js

cd apps/admin
cp .env.local.example .env.local     # fill URL + anon key
npm install
npm run dev                          # http://localhost:3000

Deploy free on Vercel: import the repo, set Root Directory = apps/admin, add the two NEXT_PUBLIC_SUPABASE_* env vars. See DEPLOY.md.

3) Mobile — Expo (APK, no Android Studio)

cd apps/mobile
cp .env.example .env                 # fill URL + anon key
npm install
eas build --platform android --profile preview   # → downloadable .apk

Details in apps/mobile/BUILD_APK.md. Face recognition requires a dev build — see apps/mobile/FACE_RECOGNITION.md.

📖 How to Use

1. Admin adds an office/post location (use "pakai lokasi saya" on-site) and registers employees (manually or via Excel import), each with an NIP.
2. Employee installs the app and logs in with their NIP (default password = NIP), then sets a new password on first login.
3. Employee taps Absen sekarang → the app checks distance to the location (Haversine), detects fake GPS, takes a proof photo, and records the check-in/out.
4. The check-in appears in the admin Reports page, ready to export as CSV.

Data out of sync after install? See apps/mobile/TROUBLESHOOTING.md — usually the Confirm email toggle or a missing migration.

🗺️ Roadmap

• Phase 0 ✅ — location + photo attendance, multi-tenant, web admin, co-brand.
• Phase 1 🚧 — face recognition (ML Kit liveness + MobileFaceNet); math + DB columns ready, embedding extraction needs a dev build.
• Phase 2 ✅ — shifts, schedules (bulk assign), Excel import, reports + CSV.
• Phase 3 ✅ — super admin: create/manage institutions with no redeploy.
• Phase 4 ⏳ — FCM notifications, MapLibre map picker, anti-fraud hardening.

🔒 Privacy

Muster stores location and biometric (face) data. Handle it under Indonesia's PDP law: state the purpose of collection, store only what's needed (embeddings, not raw photos where possible), and provide access/erasure. This is not legal advice.

🤝 Contributing

Issues and PRs welcome. The backend is plain PostgreSQL + RLS; the apps are standard Next.js and Expo — add a page or a screen and wire it to Supabase.

⚠️ Disclaimer

For legitimate workforce attendance on systems and staff you are authorized to manage. The author is not responsible for misuse.

📄 License

MIT © 2026 Abil Khosim.

---

👤 Developed by Abil Khosim

Cybersecurity Specialist

Muster is an original project by Abil Khosim, an independent security tool by Abil Khosim (NoxNull). Released under the MIT License — © 2026 Abil Khosim. Please keep this attribution when reusing or redistributing.

_{Everyone present. Everyone accounted for. 📋}