This release includes 1 security fix for security teams reviewing exposed deployments.
Topics
+4 more
Summary
AI summaryPreserve Gemini 3 thought signatures across function‑call replays and add explicit prompt cache breakpoints for GPT-5.6 models.
Full changelog
[1.12.1] - 2026-07-22
Added
- agent-framework-openai: Add explicit prompt cache breakpoints for GPT-5.6 models and a usage sample (#7163)
Changed
- agent-framework-ag-ui: Promote the package from release candidate to stable
- agent-framework-core: Add security guidance for custom MCP Streamable HTTP clients (#7245)
Fixed
- agent-framework-gemini: Preserve Gemini 3 thought signatures across function-call replays (#7095)
- agent-framework-core, agent-framework-foundry, agent-framework-foundry-hosting, agent-framework-openai: Fix stateless replay of reasoning-paired tool calls (#7233)
Full Changelog: https://github.com/microsoft/agent-framework/compare/python-1.12.0...python-1.12.1
Security Fixes
- agent-framework-core: added security guidance for custom MCP Streamable HTTP clients
Weekly OSS security release digest.
The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.
No spam, unsubscribe anytime.
Share this release
About agent-framework
A framework for building, orchestrating and deploying AI agents and multi-agent workflows with support for Python and .NET.
Related context
Related tools
Earlier breaking changes
- vpython-1.8.0 [BREAKING — experimental] Refactor Skill API to async resource and script lookup
- vpython-1.8.0 [BREAKING] Upgrade github-copilot-sdk to v1.0.0 (stable)
- vdotnet-1.9.0 Removes [Experimental] tag from .NET Orchestrations, marking them stable.
- vpython-1.7.0 Remove Python-only declarative actions and rename alias kinds to C# canonical names in agent-framework-declarative.
- vpython-1.6.0 Enable instrumentation by default in agent-framework-core and agent-framework-foundry.
Beta — feedback welcome: [email protected]