Lesson 1.2Lesson 2 of 30
Coordinator and subagents
Orchestrate multi-agent systems with coordinator-subagent patterns
Key points
- 1Hub-and-spoke: the coordinator decomposes the task, delegates, aggregates and handles errors. Subagents never talk to each other directly.
- 2Routing everything through the coordinator gives one place for observability and consistent error handling.
- 3The coordinator picks which subagents a query needs; a simple question should not run the whole pipeline.
- 4Split scope by distinct subtopics or source types. Too narrow a split leaves parts of a broad topic uncovered.
- 5Iterate: check the synthesis for gaps, send targeted follow-up queries, re-run synthesis.
What the exam guide checks here
Objective wording from the exam guide (English).
- 1.2.K1Hub-and-spoke: coordinator manages all inter-subagent communication, error handling, information routing
- 1.2.K2Subagents have isolated context and do not inherit the coordinator's history
- 1.2.K3Coordinator role: decomposition, delegation, aggregation, choosing subagents by query complexity
- 1.2.K4Risk of overly narrow decomposition leading to incomplete coverage of broad topics
- 1.2.S1Coordinator dynamically selects which subagents to invoke instead of always running the full pipeline
- 1.2.S2Partition scope across subagents (distinct subtopics or source types) to minimise duplication
- 1.2.S3Iterative refinement: coordinator finds gaps in synthesis, re-delegates targeted queries, re-invokes synthesis
- 1.2.S4Route all subagent communication through the coordinator for observability and consistent error handling
Lock it in
Answer scenario questions on this topic. Every option comes with an explanation.
Practise 1.2 · 23 questions