Free & Open — Sovereign Tools
▌▌━━🐉 ⚔️ 🛡️ 🖖 🛸 ⚖️ 👑 ⚜️ 🔮 ✨ 🏰 ━━━━▌▌▌

Download KoRT Tools

Everything you need for sovereign AI infrastructure. Free, open-source, and built on bare metal — no commercial clouds.

Gatekeeper v4.0 — Online
KoRT-Code v1.0 — Available
Lady of the Lake — Active
Offline Node — Available
⭐ Featured Release

⚔️ KoRT-Code

v1.0.0 · Based on OpenCode · MIT License

A sovereign AI coding terminal built for power users. Routes through the KoRT Gatekeeper's multi-model pipeline — Flash pre-processor, heavy lifter, reviewer — all on your own infrastructure.

3-stage multi-model pipeline (Flash → Pro → Reviewer)
KoRT Gatekeeper as primary provider (100% sovereign)
Lady of the Lake memory & SOP system built-in
Merlin, Knight, Gatekeeper agents pre-configured
Web UI on port 4096 + CLI mode
Docker-compose one-liner deploy
⬇ Download App ⭐ KoRT Git 📖 Install Guide
KoRT-Code Terminal — v1.0.0
# One-liner install $ curl -fsSL https://kortx.ca/install.sh | bash ▌▌━━🐉 KoRT-Code Installer v1.0 ━━🐉━━▌▌ ✓ Checking system requirements... ✓ Node.js 20.x — found ✓ Installing OpenCode base... ✓ Applying KoRT branding & config... ✓ Connecting to Gatekeeper on :8095 ✓ Lady of the Lake memory — initialized 🔮 Pipeline: Flash→Pro→Reviewer active ✓ KoRT-Code ready on http://localhost:4096 $
🔮 All Downloads

Sovereign Infrastructure Stack

🔮
KoRT Gatekeeper
v4.0.0 · FastAPI · Python 3.11+
Multi-model routing API. 3-stage pipeline (Flash→Pro→Reviewer). Routes to Vultr inference endpoints. Drop-in OpenAI-compatible API on port 8095.
# Docker (recommended) docker pull kortx/gatekeeper:4.0 docker run -p 8095:8095 kortx/gatekeeper:4.0 # From source git clone https://git.kortx.ca/gatekeeper pip install -r requirements.txt python main.py
→ View on KoRT Git
🏰
Lady of the Lake
v2.0.0 · Python · Offline-capable
AI learning agent & memory guardian. Analyzes failures, generates SOPs, refines routing rules. Offline-capable SQLite backend with 5 pre-seeded SOP categories.
# Run learning cycle python lady_of_the_lake.py cycle # Query wisdom python lady_of_the_lake.py wisdom timeout # Seed SOP library python lady_of_the_lake.py seed
→ View on KoRT Git
📡
Gatekeeper Offline
v2.0.0 · FastAPI port 8096
Full offline fallback node. Cached responses, memory CRUD, SOP library, model performance tracking, and sync endpoint for when connectivity is restored.
# Start offline node python offline_server.py # Sync to main gatekeeper curl -X POST http://localhost:8096/v1/sync # Check SOPs curl http://localhost:8096/v1/sops
→ View on KoRT Git
🏰
KoRT Portal
v1.0.0 · React 18 · Vite
Sovereign web portal with 10+ integrated apps: Dashboard, Knight Dispatch, Agent Dossiers, Games, Quartermaster, SalesWizard, and more.
# Self-host the portal git clone https://git.kortx.ca/portal npm install npm run build # Serve dist/ folder npx serve dist -p 3000
→ View on KoRT Git
🖥️
KoRT OS Server Panel
v1.0.0 · React + SSH
Browser-based server management for KoRT infrastructure. Monitor VMs, restart services, tail logs, and run commands from the portal.
# Part of the Portal app # Access at: /portal/server # Or standalone git clone https://git.kortx.ca/kort-os npm install && npm run dev
→ View on KoRT Git
🐋
KoRT Full Stack
docker-compose · All services
Single docker-compose file brings up the entire KoRT stack: Gatekeeper, Offline Node, Portal, and KoRT-Code web interface. Production-ready.
# Full sovereign stack curl -O https://kortx.ca/docker-compose.yml docker-compose up -d # Services: # :8095 Gatekeeper # :8096 Offline Node # :4096 KoRT-Code UI
→ Get docker-compose.yml
📖 Installation

Quick Start Guide

⚔️ KoRT-Code Setup

1

Prerequisites

Install Node.js 20+, Python 3.11+, Docker, and git. All modern Linux distros and macOS work. Windows via WSL2.

2

Clone & Configure

Run git clone https://git.kortx.ca/kort-code and copy config/global/opencode.json to ~/.config/opencode/config.json.

3

Start Gatekeeper First

KoRT-Code connects to Gatekeeper on :8095. Run python gatekeeper/main.py or docker-compose up gatekeeper first.

4

Launch KoRT-Code

Run bash scripts/deploy-all.sh or npm install -g opencode-ai && opencode. Web UI opens on http://localhost:4096.

🔮 Gatekeeper + Lady of the Lake

1

Install Dependencies

In gatekeeper/: pip install fastapi uvicorn httpx pydantic. Or use the included requirements.txt.

2

Start Gatekeeper

Run python main.py. Gatekeeper starts on :8095 with the full 3-stage pipeline. Check /health and /v1/pipeline/status.

3

Seed the SOP Library

In gatekeeper-offline/: python lady_of_the_lake.py seed. Loads 5 proven SOPs covering timeouts, rate limits, deps, permissions, networking.

4

Run a Learning Cycle

After some usage: python lady_of_the_lake.py cycle. Analyzes failures, generates new SOPs, suggests routing improvements, expires stale memories.

System Requirements

💻
OS
Linux / macOS / WSL2
🟢
Node.js
v20+ (LTS recommended)
🐍
Python
3.11+ required
🐋
Docker
Optional (compose deploy)