Beyond One Agent - Orchestrating Claude Multi-Agent Systems in Make Grid

​A smaller, deeper session for people already building with Claude and hitting the hard part: what happens when one agent isn't enough.

​

​We'll get into orchestration — designing specialised Claude subagents, delegating across them, managing context windows and handoffs, and handling errors without silent failures. Then observability: how Make Grid gives you a single view of your Make automations, Make agents and Claude Managed Agents, so you can trace a multi-agent run end to end and find where it actually broke.

​

​We'll cover:

• ​Designing Claude agent teams and delegating subtasks on Make's execution canvas

• ​Context-window management, prompt versioning and authentication patterns that survive production

• ​Human-in-the-loop checkpoints for high-stakes decisions

• ​Monitoring, tracing and debugging Claude Managed Agents in Make Grid

​Format: small group, open discussion. Bring a real architecture problem — we'll work through them together.

​

​Who this is for: Claude users already running agents, or close to it. If you're starting from zero, join Monday's hands-on workshop instead.

​

​Every attendee receives 25,000 Make credits.

Full Claude Community House on claudebcn.com