Give a complex multi-step task — the orchestrator decomposes it, picks the right agents, runs them in pipeline, and synthesizes a unified output.
POST /api/agents/orchestrator · Method: tasks/send or tasks/sendSubscribe · Agent card: GET /api/agents/orchestrator
An LLM planner analyzes your request and outputs a JSON plan. The plan executes through the A2A protocol, with outputs piped between agents.
Up to 5 sequential steps. For longer chains, run multiple orchestrator calls.
Currently auto-planned. A future version will let you specify the pipeline yourself.