C.O.R.E Ecosystem
The Cognitive Operational Response Engine. A JARVIS-like personal AI ecosystem that listens, plans, and executes — across your desktop, your phone, and your entire workflow.
One mind. Three bodies.
C.O.R.E is architected as three coordinated subsystems — each running independently, all speaking the same language.
CORE_Background
The Sleeper Agent
Runs invisibly from startup. Listens 24/7 for your wake word in English (offline Vosk) or Arabic (cloud). Hosts the mobile server and monitors your network — silently.
CORE_Desktop
The Command Center
A glassmorphism Chromium dashboard. Live system metrics, process and port managers, network security, AI chat, configurable quick actions — your entire PC, on one pane.
CORE_Ultra
Agentic Execution
Gemini 3 Flash with native audio. Plans, executes, and verifies multi-step tasks across 17 action modules — and remembers you between sessions.
Built to act, not just answer.
Every capability designed for the way power users actually work.
Offline English. Cloud Arabic.
Vosk + Kaldi closed-grammar STT for English with zero false positives. Google Speech for natural-language Arabic. Two engines, one assistant.
Learn more17 autonomous actions. Powered by Gemini 3 Flash and 30 OpenRouter models.
A full-screen agentic overlay with native audio streaming. Plans, executes, and self-heals multi-step tasks across your system.
Learn more
Control your PC from your phone.
A native Android companion that pairs over Wi-Fi via QR code. Voice commands, quick actions, and a full virtual remote — all TLS-encrypted, local-only.
Learn more
ARP monitoring built-in.
Continuous local-network scanning, MAC vendor lookup, unknown-device alerts, and per-IP ARP blocking. No third-party tools required.
Learn more
C.O.R.E remembers you.
Facts extracted from every conversation are stored, ranked, and re-injected into future sessions. Identity, projects, preferences — all preserved.
Learn moreStudy. Dev. Break. Gaming. On command.
Voice-triggered workspaces that reconfigure your environment — virtual desktops, folder shortcuts, focus layouts — instantly.
Learn moreEnglish and Arabic, natively.
Microsoft David, Zira, Naayf (Saudi), and Hoda (Egyptian) voices. Switch languages mid-session and TTS adapts in real time.
Learn moreYour voice. Your commands. Your config.
Unlimited custom voice commands, configurable wake words, AI personality, shutdown PIN, mode triggers — every detail under your control.
Learn more
1 7 plug-in capabilities.
Gemini invokes any of these autonomously to complete your request — and chains them when needed.
Your PC. In your pocket.
The Android companion turns your phone into a wireless remote for your entire workstation — over the same Wi-Fi, with zero external servers.
- Pair instantly with a QR code over local Wi-Fi
- TLS-encrypted HTTPS, zero external servers
- Send voice commands processed exactly like PC input
- Fully customizable Quick Action buttons
- Bilingual on-device AI chat (EN / AR)
- Wake your PC remotely from anywhere on your network
Scan the QR from C.O.R.E Desktop with your phone.
C.O.R.E Desktop generates a local-network QR that encodes your PC's IP and a TLS certificate fingerprint. Open C.O.R.E Connect, point the camera, and you're paired — no accounts, no cloud, no exposed ports.
Every tab, a superpower.
Offline voice commands
Tap the mic and speak — your phone streams the audio straight to C.O.R.E on your PC. Same offline English engine, no cloud round-trip.
Bilingual chat with Run-on-PC
Ask in English or Arabic, get code blocks back, and hit Run on PC to execute the snippet directly inside your workstation.
Clipboard & file exchange
Sync clipboard text both ways, push files into your PC's drop folder, and browse what's there — all over your local Wi-Fi only.
Get C.O.R.E.
Two editions. One ecosystem. Pick your platform.
Windows Edition
Windows 10/11 Desktop App + Android Companion
- Windows 10/11
- Python 3.10+ (bundled)
- Npcap (auto-installed)
- 4 GB RAM minimum
Linux Edition
Debian & Red Hat based (Wayland & X11) + Android Companion
- Debian / Ubuntu (.deb) & Fedora / RHEL (.rpm)
- Wayland & X11 Display Sessions
- Verified on GNOME & Cinnamon DE (Only)
- Python 3.10+ & PulseAudio / PipeWire
Questions, answered.
No — English voice recognition runs 100% offline via Vosk + Kaldi. Only Arabic input and optional AI chat use cloud services, and they're entirely opt-in.