← Facundo

Ongoing

Agent Orchestration

A series of experiments to achieve real orchestration between agents, using my own orchestration library


This lab is about the orchestration itself, not any single product built on top of it. The recurring question across runs is a three-layer shape: a top-level director creates tasks, those tasks start worker agents when no agent is assigned, and the workers produce separate contributions toward one shared goal.

Each run below is a record of one such test — what held up, what broke, and what that implies about the coordination layer needed around the agents.

Run