Descriptions:
Bart Slodyczka delivers a full walkthrough of Claude Managed Agents — a feature in the Anthropic console (platform.claude.com) that lets developers create persistent AI agents with built-in session management, environment scoping, and a credentials vault. The video covers every stage from initial agent creation to production deployment using N8N as the integration layer.
The working example is a customer support chatbot for a home services business that answers FAQ questions, handles troubleshooting, and automatically escalates unresolved issues by creating tasks in ClickUp via OAuth. The two-part pricing model is explained clearly: standard API token consumption plus a $0.08/session-hour runtime fee charged only while the agent is actively processing — idle time between messages incurs no cost.
Key architectural concepts covered include sessions (each customer interaction is isolated with its own private context), environments (configuration namespaces shared across or scoped to individual agents), and the credentials vault for storing API keys and OAuth tokens securely. The default model in the console is confirmed as Claude Sonnet 4.6. For teams weighing Claude Managed Agents against building their own API integration, this tutorial provides a detailed, reproducible reference for the full setup and deployment cycle, including the N8N nodes required to stitch the agent into a live web chat widget.
📺 Source: Bart Slodyczka · Published April 09, 2026
🏷️ Format: Tutorial Demo
