Terminalus
Multi-agent mission control for orchestrating parallel AI coding sessions. Type one prompt, deploy a fleet of agents, ship anything. Native macOS app built with Tauri, Rust, and React.
Parallel Agent Execution
Run 1–12+ AI agents simultaneously in an auto-sizing grid. Every agent streams output in real time so you see progress as it happens.
Git Worktree Isolation
Each agent operates in its own isolated git branch. Zero merge conflicts by design — branches auto-merge back to main on completion.
Multi-Provider LLM Support
Supports Claude, OpenAI, Gemini, Ollama, Codex, and more. The Dynamic Model Optimizer routes each task to the best model for the job.
Permission Control System
Five permission modes from plan-first to full bypass. A hold-to-confirm safety UX prevents accidental destructive operations.
Real-Time Streaming
NDJSON-based live streaming from every agent with structured output — text, tool calls, results, and errors with per-agent status indicators.
Intelligent Orchestrator
Analyzes your mission prompt, decomposes it into phased tasks, and auto-selects the right agents from your local library.