Maestro - Unified Multi-Agent AI Flow Orchestrator

Maestro is a visual dashboard for building, compiling and executing cognitive multi-agent DAG workflows, connecting a React and TypeScript canvas workspace to a Python FastAPI server and a Langflow core engine that coordinates role-playing LLM agents with real-time budget, latency and security logging. The workspace uses a fluid grid so the control sidebar, flow canvas, execution debugger console and agent settings inspector can each be resized, styled with a glassmorphic slate palette, custom micro-animations and full light and dark themes. Active-run safeguards lock composer input during execution, gate prompt edits behind a confirmation modal, and raise a global cancel event that halts timers and API polling immediately. A dynamic verification score renders as a live rating gauge that scales with workflow success and drops to zero with warning badges on failure or block, and a smart offline simulator with interactive mock generators keeps the system demonstrable without live providers.