Engineering Announcement
it’s stateless and much simpler to serve, plus a bunch of other improvements.
incredibly proud of the work our team put in alongside many other members of the community.
The post is mostly accurate, with one small nuance.
Fact check
"today marks the biggest upgrade to the mcp spec since its initial release!"
✅ Accurate.
The official MCP maintainers describe the 2026-07-28 specification as "the largest revision of the protocol since launch." (Model Context Protocol Blog)
"it's stateless and much simpler to serve"
✅ Accurate.
The new spec removes the protocol-level session and initialization handshake, making MCP stateless at the protocol layer. This allows deployments behind standard HTTP load balancers without sticky sessions or shared session stores, making servers much easier to scale and operate. (Model Context Protocol Blog)
"plus a bunch of other improvements."
✅ Accurate, though vague.
Other major additions include:
Extensions as first-class features (including MCP Apps and Tasks)
Authorization hardening aligned with OAuth/OpenID Connect
Formal deprecation policy
Full JSON Schema 2020-12 support for tools
Other protocol and SDK improvements (Model Context Protocol Blog)
"incredibly proud of the work our team put in alongside many other members of the community."
✅ Plausible, assuming the author actually contributed. The MCP spec is developed by maintainers and the broader open-source community, and the official announcement thanks contributors from the working groups and community. (Model Context Protocol Blog)
Overall verdict
🟢 Accurate (9.5/10).
The only tiny nitpick is that "much simpler to serve" is informal wording. A slightly more precise phrasing would be:
"Today marks the biggest upgrade to the MCP spec since its initial release. It's now stateless, making MCP servers much easier to deploy and scale, along with a host of other improvements."
The original post is factually sound overall.