AI Workshop
Episode 8 of 10Upcoming
A fleet of agents
Subagents, background agents and orchestration: when one agent isn't enough, and how not to drown in ten.
The link arrives by email an hour before the stream
Sign up — a day before we'll send a reminder with the episode plan, and an hour before the start the stream link. The player appears on this page 15 minutes before the start.
What we'll show
- Subagents: a separate context for search, review, tests — and why that saves the main context
- Background agents: the task runs while you do something else
- Orchestration: one agent hands out tasks, others do them, a third checks
- Where parallelism helps and where agents get in each other's way (shared files, conflicts)
- Cost: counting tokens and not burning the budget on a fleet
What you take away
An «orchestrator + workers + reviewer» layout for a task of your own.
Fleet layout and orchestration rules
- Split by independence: run in parallel only tasks that don't touch the same files
- Each agent has one role and a clear output: a report, a diff, a verdict
- The reviewer isn't the author: review by a separate agent or a human
- The orchestrator holds the overall plan; workers get a narrow context
- Background tasks end with a report you can read in a minute
- There's a token budget and a cap on concurrent agents
Harness in this episode
In a fleet the harness becomes mandatory: without shared checks and file ownership, agents break each other's work.
The full harness episodeWhat to have ready
- A task that splits into 3–4 independent parts