Artificial intelligence news.
Straight from the original announcements.

Pydantic maintainer wins AAIF award for MCP Python SDK work

October 27, 2026, 10:00Source: Pydantic
Pydantic maintainer wins AAIF award for MCP Python SDK work
Image: Pydantic

A Pydantic developer who maintains the MCP Python SDK has received the Outstanding Technical Contribution Award from the Agentic AI Foundation. The award was presented on day two of AGNTCon + MCPCon North America and recognizes work centered on the SDK's v2 release.

Key points
  • The Agentic AI Foundation gave its Outstanding Technical Contribution Award to a maintainer of the MCP Python SDK at AGNTCon + MCPCon North America.
  • v2 of the SDK is now the stable release line, built alongside the 2026-07-28 MCP specification revision.
  • v2 adds a new MCPServer API, multi-round-trip requests and OpenTelemetry tracing on by default.
  • The author also sponsored SEP-1288, the WebSocket transport, and maintains Starlette and Uvicorn.
  • The Logfire MCP server runs on the same SDK and receives these changes first.

An award for the Python SDK behind MCP servers

The Agentic AI Foundation gave the award at AGNTCon + MCPCon North America, according to a post on Pydantic's site written by the recipient. The moment can be watched on the day two livestream, right after the opening keynote.

The recipient is one of two maintainers of the MCP Python SDK, which Python developers use to build MCP servers and clients. According to the post, the SDK has taken up much of their time this year. A colleague, Laís Carvalho, nominated them.

The author stresses how much of the work is invisible to a conference audience: "Most of that work happens in issues, pull requests, and spec discussions that nobody sees on a conference stage."

What v2 brings

The largest piece of the work was v2. The team shipped the v2 beta alongside the 2026-07-28 revision of the MCP specification, with a new MCPServer API and multi-round-trip requests. According to the post, v2 is now the stable release line.

Much of v2 is the new specification implemented in Python. The post lists these items:

  • standard error codes
  • authorization hardening
  • routable transport headers
  • a deprecation policy
  • OpenTelemetry tracing on by default

On the specification side, the author sponsored SEP-1288, the WebSocket transport. The SDK's HTTP transport also runs on Starlette and Uvicorn, both of which the author maintains. That means that when the SDK exposes a bug in those layers, it can be fixed where it belongs.

Logfire integration and what it means for users

The author also works on Pydantic Logfire, so v2 ships an OpenTelemetry middleware that is enabled by default. According to the post, developers call logfire.configure(distributed_tracing=True) in their MCP server and logfire.configure() in their client. Every request between the two then arrives in Logfire as one trace. The v2 beta post walks through the setup.

The Logfire MCP server runs on the same Python SDK, so the standard error codes, authorization hardening and tracing middleware reach it before anywhere else. The author says they want it to be the best MCP server to connect a coding agent to. Users with a Logfire project can add https://logfire-us.pydantic.dev/mcp, or the EU URL, to their agent and ask what broke in production the night before. The author says setup takes a few minutes and the answer usually takes less.

The author thanked the MCP maintainers and contributors, saying the SDK is better because of every issue, review and pull request they sent. They also thanked Pydantic for paying them to work on open source and letting them spend that time where the community needs it, which they call "rarer than it should be." They shared the stage with Antoni Kozelski and Bartosz Gonczarek of Vstorm, who won the Open Agentic AI Innovation Award.

For developers building on MCP in Python, the post confirms that v2 is the stable line and that it follows the current specification revision.

Related news

Mozilla AI launches Otari Subscriptions for open-weight coding models
Products and toolsNovember 6, 2026, 18:31

Mozilla AI launches Otari Subscriptions for open-weight coding models

Mozilla AI is opening early access to Otari Subscriptions, a service designed to make open-weight models easier to use for coding agents. The plans combine multiple…

AI21 Labs launches Teel, a Python runtime for faster agent workflows
For developersNovember 5, 2026, 07:57

AI21 Labs launches Teel, a Python runtime for faster agent workflows

AI21 Labs has introduced Teel, an open source Python runtime designed to overlap independent, IO-bound calls in ordinary sequential Python code. The company says Teel…

Half of Slack messages now come from apps and AI agents
Products and toolsNovember 3, 2026, 23:30

Half of Slack messages now come from apps and AI agents

Half of Slack messages are now sent by apps and AI agents, Salesforce said. The platform processes 127 million automated MCP operations each week, a figure that doubles…