2 Releases published by 1 person

• python-v0.5.4

  published Apr 22, 2025

• python-v0.5.5

  published Apr 25, 2025

32 Pull requests merged by 11 people

• Add more mcp workbench examples to MCP API doc

  #6403 merged Apr 26, 2025

• Update website for v0.5.5

  #6401 merged Apr 25, 2025

• Document custom message types in teams API docs

  #6400 merged Apr 25, 2025

• [doc] Clarify selector prompt for SelectorGroupChat

  #6399 merged Apr 25, 2025

• Add functional termination condition

  #6398 merged Apr 25, 2025

• Update: implement return_value_as_string for McpToolAdapter

  #6380 merged Apr 25, 2025

• Update version to 0.5.5

  #6397 merged Apr 25, 2025

• Add example using autogen-core and FastAPI for handoff multi-agent design pattern with streaming and UI

  #6391 merged Apr 25, 2025

• Update agent documentation

  #6394 merged Apr 25, 2025

• AssistantAgent to support Workbench

  #6393 merged Apr 24, 2025

• TEST: skip when macos+uv and adding uv venv tests

  #6387 merged Apr 24, 2025

• Add guide for workbench and mcp & bug fixes for create_mcp_server_session

  #6392 merged Apr 24, 2025

• Remove name field from OpenAI Assistant Message

  #6388 merged Apr 24, 2025

• Introduce workbench

  #6340 merged Apr 24, 2025

• Revert "Update: override return_value_as_string for McpToolAdapter"

  #6378 merged Apr 24, 2025

• Update: override return_value_as_string for McpToolAdapter

  #6371 merged Apr 24, 2025

• TEST/change gpt4, gpt4o serise to gpt4.1nano

  #6375 merged Apr 23, 2025

• FEAT: MCP Actor for WorkBench for #6340

  #6360 merged Apr 22, 2025

• update website version

  #6364 merged Apr 22, 2025

• Add self-debugging loop to CodeExecutionAgent

  #6306 merged Apr 22, 2025

• Fix: deserialize model_context in AssistantAgent and SocietyOfMindAgent and CodeExecutorAgent

  #6337 merged Apr 22, 2025

• Add azure ai agent

  #6191 merged Apr 22, 2025

• Avoid re-registering a message type already registered

  #6354 merged Apr 22, 2025

• Added support for exposing GPUs to docker code executor

  #6339 merged Apr 22, 2025

• fix: ollama fails when tools use optional args

  #6343 merged Apr 22, 2025

• Add an example using autogen-core and FastAPI to create streaming responses

  #6335 merged Apr 21, 2025

• FEAT: SelectorGroupChat could using stream inner select_prompt

  #6286 merged Apr 21, 2025

• Add experimental notice to canvas

  #6349 merged Apr 21, 2025

• DOC: add extentions - autogen-oaiapi and autogen-contextplus

  #6338 merged Apr 21, 2025

• fix: ensure serialized messages are passed to LLMStreamStartEvent

  #6344 merged Apr 21, 2025

• Generalize Continuous SystemMessage merging via model_info[“multiple_system_messages”] instead of startswith("gemini-")

  #6345 merged Apr 21, 2025

• Agentchat canvas

  #6215 merged Apr 21, 2025

7 Pull requests opened by 7 people

• feat: Add MessageStore base class abstraction for storing message thread in teams; Implement MemoryMessageStore for internal uses

  #6350 opened Apr 21, 2025

• feat: Add id field for AgentChat messages

  #6351 opened Apr 21, 2025

• [Experimental] HTTP agent runtime

  #6356 opened Apr 22, 2025

• Fix: Icons are not aligned vertically.

  #6369 opened Apr 23, 2025

• Bing grounding citations

  #6370 opened Apr 23, 2025

• Serialization for the functions, fixed issues in agentchat selector_func and candidate_func serialization

  #6389 opened Apr 24, 2025

• [FIX] DockerCommandLineCodeExecutor multi event loop aware

  #6402 opened Apr 26, 2025

11 Issues closed by 3 people

• SelectorGroupChat needs for clarification for metadata

  #6383 closed Apr 25, 2025

• OperationalError: no such column: "size" - should this be a string literal in single-quotes?

  #6009 closed Apr 25, 2025

• macOS/uv: venv tests fail (dyld error), not resolved by symlinks=False

  #6341 closed Apr 24, 2025

• Error in SseMcpToolAdapter.from_server_params shows Server error '502 Bad Gateway' for url

  #6382 closed Apr 24, 2025

• An error occurred: Error code: 500 - {'error': {'message': 'The model produced invalid content. Consider modifying your prompt if you are seeing this error persistently.', 'type': 'model_error', 'param': None, 'code': None}}[Issue]:

  #3247 closed Apr 24, 2025

• Self-debugging in `CodeExecutionAgent`

  #6207 closed Apr 22, 2025

• Some agents do not deserialize model_context (set to None or Do not serialize too)

  #6336 closed Apr 22, 2025

• Teams with agents that contain duplicated produced_message_types result in an ValueError exception

  #6353 closed Apr 22, 2025

• Add optional `device_requests` to DockerCommandLineCodeExecutor

  #6302 closed Apr 22, 2025

• Support anyOf property types in convert_tools for Ollama clients

  #6323 closed Apr 22, 2025

• Model client streaming from the selector of SelectorGroupChat

  #6145 closed Apr 21, 2025

11 Issues opened by 10 people

• Function call with Gemini have no code samples

  #6396 opened Apr 25, 2025

• DockerCommandLineCodeExecutor does not safely manage cancellation tasks across multiple event loops (loop mismatch issue)

  #6395 opened Apr 25, 2025

• Error in LangChainToolAdapter GoogleDriveSearchTool: Unable to generate pydantic-core schema

  #6385 opened Apr 24, 2025

• `ThoughEvent` triggers termination via `TextMentionTermination`

  #6384 opened Apr 24, 2025

• Tool id string too long

  #6377 opened Apr 23, 2025

• autogen-ext test is too slow...!

  #6376 opened Apr 23, 2025

• Error in sse_reader for MCP

  #6372 opened Apr 23, 2025

• Update `McpToolAdapter` to implement `return_value_as_string`

  #6368 opened Apr 23, 2025

• The icons are not aligned vertically.

  #6366 opened Apr 23, 2025

• **Problem Statement** Authorization specification for AutoGen clients

  #6355 opened Apr 22, 2025

• completion_tokens return None when result.usage is None

  #6352 opened Apr 21, 2025

13 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Added Graph Based Execution functionality to Autogen

  #6333 commented on Apr 23, 2025 • 12 new comments

• Add `model_context` to `SelectorGroupChat` for enhanced speaker selection

  #6330 commented on Apr 23, 2025 • 10 new comments

• Facing Issue while working with Groq openAI compatible endpoint

  #6255 commented on Apr 19, 2025 • 0 new comments

• AgentChat messages should have an `id` field.

  #6317 commented on Apr 21, 2025 • 0 new comments

• Make usage of `name` field in OpenAI messages optional in `OpenAIChatCompletionClient`

  #6034 commented on Apr 21, 2025 • 0 new comments

• Introduce `OpenAIAgent` backed by the Response API in Extensions.

  #6032 commented on Apr 23, 2025 • 0 new comments

• Autogen website accessibility issues (low severity)

  #6090 commented on Apr 24, 2025 • 0 new comments

• [Issue]: The problem of downloading monaco-editor when accessing autogenstudio in offline environment.

  #3556 commented on Apr 25, 2025 • 0 new comments

• Adding bedrock chat completion for anthropic models

  #6170 commented on Apr 22, 2025 • 0 new comments

• Implements suggestion discussed in issue #6049 by adding raw_response field to CreateResult.

  #6192 commented on Apr 25, 2025 • 0 new comments

• [FEAT] summarized chat completion context

  #6217 commented on Apr 21, 2025 • 0 new comments

• Create a common abstraction for storing message threads in teams

  #6243 commented on Apr 22, 2025 • 0 new comments

• Remove duplicate aliases

  #6309 commented on Apr 25, 2025 • 0 new comments