Agent DailyAgent Daily
releaseintermediate

[Release] microsoft/autogen python-v0.6.4: python-v0.6.4

By ekzhugithub
View original on github

Microsoft AutoGen python-v0.6.4 release includes improvements to GraphFlow for better state management and task resumption, enhancements to Workbench implementations with tool name/description overrides, and fixes for function calling support and conversation context pollution. The release also adds new features like Qwen2.5VL support, DuckDuckGo search tools, and improved documentation generation.

Key Points

  • GraphFlow now retains execution state after termination and can be resumed with new tasks without explicit reset, matching behavior of RoundRobinGroupChat and SelectorGroupChat
  • StopAgent has been removed from GraphFlow; stop messages are now carried in the TaskResult.stop_reason field instead of polluting conversation context
  • GraphFlow supports multiple sequential task execution within the same instance, with state reset only when no more agents are available
  • McpWorkbench and StaticWorkbench now allow overriding tool names and descriptions for client-side optimization and better server-side tool adaptability
  • Fixed function calling support for Llama3.3 model to ensure compatibility with the latest Llama version
  • Added Claude model reflection support in AssistantAgent for improved reasoning capabilities
  • Added Qwen2.5VL model support to expand multi-modal AI model compatibility
  • Removed otel semcov package from core dependencies to reduce dependency footprint
  • Documentation restructured by moving docs from python/packages/autogen-core to python/docs with automated API documentation generation
  • Updated GitHub Models URL and added DuckDuckGo Search Tool integration to AutoGen Extensions

Found this useful? Add it to a playbook for a step-by-step implementation guide.

Quality

Concepts