Vibeleaderboard
← All Intel
Intel / repo

semantic-kernel python-1.47.0

Source
github.com
Author
eavanvalkenburgTop Viber
Date

Context

The change Python users need to plan for in Semantic Kernel python-1.47.0 is the version floor: the release requires Python 3.11 or later, removes support for Python 3.10, and is tested on 3.11 through 3.14 (release notes). Semantic Kernel is Microsoft's open-source SDK, a code library developers use to connect their applications to AI models and build agents on top of them. A project still running on Python 3.10 has to move to a newer interpreter before it can install this version. The rest of the Python release is consolidated dependency updates and bug fixes, including fixes for the Mistral AI, Google AI, SQL Server, Redis and Oracle connectors.

Two of those fixes tighten a check in the OpenAPI plugin, the component that lets an call a web service described by an OpenAPI specification. Developers can restrict that plugin to an of base URLs that include a path, such as /api. Before this release, a path written with dot segments, like /api/../admin or its encoded form /api/%2e%2e/admin, could pass an /api allowlist while the HTTP client actually requested /admin; the fix resolves those segments before comparing (pull request 14668). A second fix stops the check from ignoring letter case, because URL paths are case-sensitive and a path that differs only in case is a different path (pull request 14693). The release notes list both as ordinary fixes, but both close ways for a request outside the allowed path to get through.

The project's README now calls Microsoft Agent Framework "the enterprise-ready successor to Semantic Kernel" and links a migration guide (repository). This release continues to ship fixes for teams that still run Semantic Kernel itself.

Terms in this piece · Glossary
  • AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
  • tool permissions — The rules governing which tools an agent may call and which need confirmation — the boundary between a mistake and an incident.
  • embedding — A list of numbers representing a piece of text's meaning, so that similar meanings end up numerically close and can be searched.
Recommended reads
Comments

Checking sign-in…

Loading comments…