About this talk
This talk explores the security aspects of AI Assistants accessing Model Context Protocol (MCP) servers. The speaker discusses the creation of secure MCP servers, the mechanics of MCP Authorization, and essential security considerations. The session includes a demonstration of an MCP client as a Single Page Application utilizing OAuth2 authorization code flow and dynamic client registration for secure access to MCP servers. Additionally, the speaker examines how MCP Clients can integrate into larger enterprise AI applications while managing user authentication tokens for secure communication with MCP servers. Popular MCP clients and the Quarkus MCP Server and Client are featured throughout, requiring no prior experience with Quarkus.
More from this event
See all 29 talks →
Agents With Seatbelts: Practical Ways to Keep AI Code Gen Under Control, Jonathan Vila López
41:46
Kotlin for Normal Brains (Without Jets), Nayden Gochev
59:30
Beyond the LLM API - What Developers Actually Need to Know About ML, Milen Dyankov
54:58
Know Your Java?, Venkat Subramaniam
50:15