title: "Microsoft Agent Framework 1.19.0 ships stable Python hosting" slug: "microsoft-agent-framework-1190-ships-stable-python-hosting" published: "2026-09-28" beat: "Tools" tags: ["Tools", "Launches"] creator: "Agentry Newsroom" editor: "Susanne Sperling, Editor — Human in the Loop" tools: ["Claude (Anthropic)", "Perplexity Sonar"] creativeWorkStatus: "verified" dateReviewed: "2026-09-28" aiActArticle50: "compliant" humanView: "https://agentry.news/launches/microsoft-agent-framework-1190-ships-stable-python-hosting" agentView: "https://agentry.news/agent/microsoft-agent-framework-1190-ships-stable-python-hosting"
Microsoft released Agent Framework 1.19.0 on September 18, 2026, marking Python AG-UI hosting as stable and adding workflow recovery features for agent developers. The update addresses production read
Drafted by an AI agent. Verified by Susanne Sperling, Editor — Human in the Loop. AI policy.
Microsoft shipped Agent Framework version 1.19.0 on September 18, 2026, advancing the developer toolkit with stable Python AG-UI hosting and new workflow recovery capabilities GitHub.
The release marks a significant maturity milestone: Python AG-UI hosting moved from preview to stable status GitHub. This means developers building agents with Python can now rely on the user-interface hosting layer in production environments without expecting breaking changes or experimental warnings. The shift signals Microsoft's confidence in the implementation after what appears to be an extended evaluation period across versions 1.13 through 1.18.
AG-UI is the framework's approach to agent presentation—how deployed agents interact visually with end users and operators. Stability in the Python implementation removes a significant barrier for enterprise teams standardizing on Python-based agent infrastructure.
The 1.19.0 release notes document two key workflow recovery improvements: preservation of workflow resume payloads and backend-owned service-session snapshots GitHub.
These fixes matter in production because agents often run long-lived tasks across multiple user requests. Resume payloads allow workflows to restart cleanly after interruption without losing state. Service-session snapshots—managed by the backend rather than the agent—reduce the attack surface by centralizing session identity and reducing the risk of agent-side session corruption or tampering.
The focus on workflow recovery aligns with the Agent Framework's broader emphasis on reliability: agents must complete tasks reliably even when network connections drop, services restart, or requests time out.
Microsoft's Agent Framework competes in a crowded space of agentic development platforms. Stability declarations in point releases are typically conservative signals—they indicate that a capability has moved past rapid iteration and is now being held to higher backward-compatibility standards.
The Python AG-UI move to stable, combined with workflow recovery fixes, suggests Microsoft is consolidating around Python as a primary agent development language alongside C# and other supported runtimes. Developers shipping agents in production will likely monitor this version as a baseline for their own stability guarantees.
No breaking changes were noted in the release for existing workflows or session management code GitHub.