MCP Release Party
The Model Context Protocol (MCP) is getting an upgrade!
Join the global community as we celebrate the latest MCP release with an evening of technical talks, community discussions, and networking. Whether you're already building with MCP or just starting to explore the ecosystem, this is an opportunity to learn what's new, meet other builders, and connect with the people shaping the future of AI interoperability.
This event is part of a series of MCP Release Parties taking place simultaneously in cities around the world.
What We'll Cover
We'll explore what's new in the latest MCP release, hear from builders and contributors across the ecosystem, and discuss how the protocol is being adopted in real-world AI applications.
Expect technical talks, community conversations, demos, and plenty of opportunities to connect with fellow developers.
Agenda
05:30 pm — Doors open & networking
06:00 pm — Welcome
06:15 pm — Keynote: Two Years of MCP
06:35 pm — Speaker Sessions
07:35 pm — Community Q&A
08:00 pm — Networking
08:30 pm — End of the event
Sessions
Keynote: Two Years of MCP
A look at what's landing in the 2026-07-28 MCP spec release, why the changes matter for people building servers and clients, and where the protocol goes from here, two years into its evolution.
About Speaker:
Den Delimarsky is a Member of Technical Staff at Anthropic and one of the lead maintainers of the Model Context Protocol.
Session 1: MCP Goes Stateless
The new MCP specification (2026-07-28) makes the protocol stateless. We will break down what this change is, why it matters, and then dig into it together with a live demo.
About Speakers:
John Dellenbaugh is a Software Engineering Manager, most recently at Microsoft, where he spent 19 years delivering cloud, security, and AI work with a hands-on background in software security, network security, and Zero Trust identity. He now works as a founding engineer at an early-stage AI startup on MCP architecture, vector search, and agent workflows.
Pankaj Kumar is a Software Developer at Amazon with a deep interest in MLOps, especially the infrastructure and emerging standards for running AI agents in production. Outside of work, he is a big fan of Lord of the Rings, anime, and manga.
Session 2: Say What You Mean: JSON Schema 2020-12 and the Contract for Context
In MCP, the schema is the contract for context. It's how a tool tells a model what goes in and what comes out. MCP originally shipped a three-keyword subset of JSON Schema, a reasonable bet on the lowest common denominator across a fragmented landscape of LLM APIs. Two SEPs later, tool schemas conform to JSON Schema 2020-12, the same default dialect as OpenAPI 3.1. This talk will cover examples of what that unlocks, what it doesn't fix, and what client and server authors need to change in the new specification.
About Speaker:
Ola Hungerford is a Principal Engineer at Nordstrom and a maintainer and community moderator for the Model Context Protocol. She leads AI enablement initiatives while contributing to MCP's specification, developer tooling, documentation, and community governance. Ola comes from a diverse technical background spanning two decades, including personalisation systems and enterprise platforms. She's also a vocalist and songwriter who experiments with augmenting human creativity through technology.
Session 3: Governed MCP: Policy Enforcement and Tamper-Evident Audit Chains
MCP servers today have no built-in way to enforce policy or prove what happened after the fact. This session walks through cMCP, an open-source gateway that wraps any MCP server without requiring changes to it, enforcing Cedar policies before each tool call and chaining every call into a tamper-evident record. We'll cover how it works, see it running against a live server, and discuss where this fits into the broader MCP ecosystem as adoption grows.
About Speaker:
Imran Siddique is Chief Platform Officer at Opaque Systems and creator of the Agent Governance Toolkit, an open-source project (github.com/microsoft/agent-governance-toolkit) with ~5k GitHub stars positioned for the Agentic AI Foundation under the Linux Foundation. He founded agentrust-io (github.com/agentrust-io), an open standards initiative behind Agent Manifest and Confidential MCP. Imran spent 18 years across various parts of Microsoft Azure, most recently leading agentic AI architecture and engineering strategy on Azure Core Engineering. He's a published author who treats governance as an engineering discipline, not a checkbox.
Session 4: Events Notifications in MCP
MCP, the protocol agents use to reach tools and data, can notify a client only while it holds a connection open, and mostly only that something changed, not what. That's about to change. An MCP working group is drafting Events, an upcoming protocol extension that lets agents subscribe to structured event streams. This talk is a brief tour of the proposal and how it is going to shape real-time events in MCP.
About Speaker:
Aman Singh is a Principal Software Engineer in the Office of Azure's CTO, where he focuses on Agentic Software Engineering and is a maintainer of Drasi, a CNCF Sandbox project for change-driven architectures. He has had more than 12 years of distributed systems experience at LinkedIn and Microsoft, spanning stream processing (Kafka, Samza), cloud migration, and building new Azure services from the ground up.
Who Should Come
AI engineers, developers, founders, researchers, open source contributors, and anyone interested in the future of AI interoperability.
Whether you're already building MCP servers and clients or simply curious about the protocol, you'll find people to learn from and connect with.
About AAIF
The Agentic AI Foundation (AAIF) is a community dedicated to advancing the understanding and practical application of agentic AI. We bring together engineers, researchers, founders, builders, and AI enthusiasts to explore how autonomous AI systems are designed, evaluated, and deployed.
Through reading groups, workshops, and community discussions, AAIF brings together engineers, researchers, founders, and AI enthusiasts to exchange ideas, challenge assumptions, and learn from one another.
By attending, you agree to our Code of Conduct and Privacy Policy.