📸 Real-Time Telemetry & Dashboard

"Visualizing complex multi-agent cognitive trees in real time."

1. High-Speed Terminal UI & Active Agents Table

Monitor 50+ agents, budget caps, and model health directly from your terminal.

2. Autonomous Communication Mesh & Real-Time Metrics

Graph-based routing between GPT-6 Astra, Claude 5.5 Opus, Gemini 4 Carbon, and DeepSeek-V4.1.

⚡ Why LEGION?

Legacy frameworks (CrewAI, AutoGen) break down when scaling beyond 5–10 agents due to state drift, token waste, and lack of real-time visibility. LEGION was built from scratch to solve the 50+ agent bottleneck.

🔥 Key Features

• ⚔️ Heterogeneous Model Mesh: Assign different LLMs to different roles in a single swarm (GPT-6 for planning, Claude 5.5 for research, DeepSeek-V4 for fast validation).
• 🛡️ Budget Guardrails & Cost Routing: Set absolute hard caps ($20.00). If a node approaches its budget limit, tasks are automatically re-routed to cheaper models (DeepSeek-V4.1-Flash).
• 🔄 Fault-Tolerant Consensus: If an agent hallucinates or times out, the Supervisor node invalidates the response and auto-respawns the sub-task.
• ⚡ Zero-Knowledge Validation: Output verification layers ensure 100% schema compliance before merging multi-agent results.
• 💻 Zero-Dependency Executable (legion.exe): Complete standalone high-performance binary with built-in rich TUI telemetry.

🚀 Quickstart

1. Download & Extract

Download the official release archive Legion.rar and unpack it to any folder.

👉 Click Here to Download LEGION v2.6.0 (Legion.rar)

2. Launch the Engine

Simply double-click legion.exe to launch the interactive TUI interface, or execute it directly from Command Prompt / PowerShell:

# Simply run the executable to start default swarm telemetry
./legion.exe

3. Advanced Configuration (Optional)

You can customize your multi-agent cluster topology by providing a legion_2026.yaml config file:

# Launch swarm execution with a custom config file and dashboard
./legion.exe run --config legion_2026.yaml --dashboard

# Run in headless mode for server deployments
./legion.exe run --config legion_2026.yaml --headless

🛠️ CLI Reference (legion.exe)

Usage: legion.exe [COMMAND] [OPTIONS]

Commands:
  run        Spawn and orchestrate an AI agent legion
  monitor    Attach real-time TUI dashboard to a running cluster session
  export     Export session logs, metrics, and state graphs (PDF/JSON)
  doctor     Verify LLM provider API keys and network latency

Options:
  --config PATH      Path to legion YAML configuration
  --all              Display all 50+ agents in terminal table view
  --port INT         WebSocket dashboard server port [default: 8080]

📅 Changelog

v2.6.0 (2026 Edition — Latest Release)

• 🚀 Added support for GPT-6 Astra, Claude 5.5 Opus, and Gemini 4 Carbon.
• ⚡ Inter-agent communication latency reduced to 140ms.
• 🛡️ Implemented Zero-Knowledge Proofs for agent task validation.
• 🎨 Complete overhaul of the Terminal UI (TUI) with real-time budget telemetry.

📜 License

Distributed under the MIT License. See LICENSE for more information.