1 Release published by 1 person

• v0.2.10

  published Aug 29, 2025

15 Pull requests merged by 7 people

• v0.2.10

  #1607 merged Aug 29, 2025

• fix: copy extra_kwargs to prevent litellm from polluting metadata

  #1610 merged Aug 29, 2025

• Fix #1564 Add conversations API support

  #1587 merged Aug 29, 2025

• Add reasoning text delta event support for gpt-oss models

  #1586 merged Aug 29, 2025

• Handle LiteLLM's json_tool_call addition for structured outputs

  #1602 merged Aug 29, 2025

• Fix #1592 Web search tool updates

  #1601 merged Aug 28, 2025

• Fix #1585 supporting connectors

  #1600 merged Aug 28, 2025

• Upgrade underlying oepnai package to the latest version

  #1599 merged Aug 28, 2025

• Remove CI job we no longer actively use

  #1590 merged Aug 26, 2025

• Update all translated document pages

  #1589 merged Aug 26, 2025

• Docs: Improve example for custom Session class implementation.

  #1582 merged Aug 26, 2025

• Feature: Make trace_include_sensitive_data configurable via environment variable

  #1192 merged Aug 26, 2025

• Fix: Emit tool_called events immediately in streaming runs

  #1300 merged Aug 25, 2025

• Update all translated document pages

  #1577 merged Aug 24, 2025

• Docs: Improvements for SQLAlchemy Sessions

  #1576 merged Aug 24, 2025

5 Pull requests opened by 5 people

• fix(responses): normalize reasoning items by inserting placeholder as…

  #1593 opened Aug 27, 2025

• [WIP] Tool guardrails

  #1606 opened Aug 28, 2025

• Enable gpt-realtime released at 2025 08 28

  #1615 opened Aug 29, 2025

• fix(run): fire on_llm_start / on_llm_end in Runner.run() for streaming & non-streaming (aligns with docs)

  #1619 opened Aug 29, 2025

• Ensure that input_guardrails can block model/tools from running

  #1622 opened Aug 30, 2025

29 Issues closed by 7 people

• Is it possible for a realtime agent to handoff to a regular agent?

  #1525 closed Aug 31, 2025

• Duplicated outputs

  #1531 closed Aug 31, 2025

• input list for agents when using as tools

  #1533 closed Aug 31, 2025

• WEB SEARCH TOOL NOT WORKING

  #1617 closed Aug 29, 2025

• Fix Agent.name Type Annotation

  #1618 closed Aug 29, 2025

• Add conversations API support

  #1564 closed Aug 29, 2025

• Add Flag for Priority Processing Tier (or Tier in general)

  #1459 closed Aug 28, 2025

• Models do not support json_schema type of response_format

  #1595 closed Aug 28, 2025

• Web search tool updates

  #1592 closed Aug 28, 2025

• Add connector support

  #1585 closed Aug 28, 2025

• GPT-5 Code Interpreter giving error with previous_response_id

  #1603 closed Aug 28, 2025

• LiteLLM + Gemini 2.5 Pro: cached_tokens=None crashes Agents SDK with Pydantic int-validation error

  #758 closed Aug 28, 2025

• Unclosed client session when using some litellm models

  #1016 closed Aug 28, 2025

• items_to_messages method not handle "ResponseReasoningItemParam"

  #1165 closed Aug 28, 2025

• Citations/Annotations Lost When Using LitellmModel with Perplexity Models

  #1346 closed Aug 28, 2025

• Issue with output_type and Custom LLM Providers in openai-agents-python

  #1390 closed Aug 28, 2025

• NotFoundError when using LiteLLM with localhost

  #1526 closed Aug 28, 2025

• Structured outputs with Gemini via LiteLLM does not work

  #1575 closed Aug 28, 2025

• Forcing MCP Use

  #1581 closed Aug 28, 2025

• Does gpt-oss support Agents SDK tools?

  #1379 closed Aug 28, 2025

• How to use ImageGenerationTool with other third-party models

  #1591 closed Aug 27, 2025

• Add support for global or environment-based default for `trace_include_sensitive_data` in `RunConfig`

  #1183 closed Aug 26, 2025

• Custom model provider agent does not use tool calling when an output type is specified

  #332 closed Aug 26, 2025

• MCP OAuth Support and Multi-Tenancy

  #1566 closed Aug 25, 2025

• SSL error for streaming voice terminal app

  #423 closed Aug 25, 2025

• How can you add session (memory) to realtime agents

  #1464 closed Aug 25, 2025

• RealtimeAgent Error(message='Conversation already has an active response in progress: Wait until the response is finished')

  #1466 closed Aug 25, 2025

• Bug: `run_streamed` Buffers `tool_call_item` Events, Delaying Real-Time Feedback

  #1282 closed Aug 25, 2025

• Streaming the final output

  #1578 closed Aug 24, 2025

10 Issues opened by 9 people

• logprobs and refusal validation error for non-OpenAI model providers

  #1616 opened Aug 29, 2025

• Support for `gpt-realtime`

  #1614 opened Aug 29, 2025

• on_llm_start / on_llm_end hooks passed to Runner aren't triggered

  #1612 opened Aug 29, 2025

• Bad Request 400 : Reasoning model issue while store is True

  #1611 opened Aug 29, 2025

• Feature Request: Support Custom Headers for MCP Servers in OpenAI Realtime API

  #1609 opened Aug 29, 2025

• Feature Request: Support for Remote Streamable HTTP MCP Servers in OpenAI Realtime API

  #1608 opened Aug 29, 2025

• Tool results as model input fails with harmony backed models (gpt-oss)

  #1605 opened Aug 28, 2025

• Configure realtime agents with Azure OpenAI's Realtime API

  #1594 opened Aug 27, 2025

• Disable SSL verification for local mcp servers

  #1583 opened Aug 25, 2025

• `tool_output` stream events don't distinguish between success and error states

  #1579 opened Aug 24, 2025

22 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.

• Save session on turn rather than at final response

  #1550 commented on Aug 29, 2025 • 3 new comments

• feat(run): lifecycle interrupt + live input injection; cancel-aware tools; safer tracing

  #1557 commented on Aug 26, 2025 • 0 new comments

• fix(voice/stt): accept conversation.item.input_audio_transcription.completed (keep legacy alias)

  #1537 commented on Aug 28, 2025 • 0 new comments

• Feat: Add Optional Structured Session Metadata

  #1474 commented on Aug 27, 2025 • 0 new comments

• Allow to pass both session and input list

  #1298 commented on Aug 27, 2025 • 0 new comments

• Enh - Use handoff_description as Default Fallback for Tool Description in as_tool() of agents

  #1199 commented on Aug 25, 2025 • 0 new comments

• feat: #864 support streaming nested tool events in Agent.as_tool

  #1057 commented on Aug 30, 2025 • 0 new comments

• Add run configuration inheritance for agent-as-tool workflows

  #1006 commented on Aug 27, 2025 • 0 new comments

• Ensure that input_guardrails can block tools from running

  #931 commented on Aug 30, 2025 • 0 new comments

• Sometimes Gemini via LiteLLM will return a response with no `choice`s

  #1559 commented on Aug 31, 2025 • 0 new comments

• Add Support for Image Return in Agent Tools

  #341 commented on Aug 30, 2025 • 0 new comments

• No OpenAI tracing with Azure Foundry models

  #1051 commented on Aug 30, 2025 • 0 new comments

• openai.BadRequestError with gpt-oss-120b - Expected 2 output messages (reasoning and final), but got 3.

  #1468 commented on Aug 30, 2025 • 0 new comments

• Context window management on long sessions

  #1539 commented on Aug 30, 2025 • 0 new comments

• Input guardrail tripwire allows tool execution to continue after exception is raised

  #991 commented on Aug 29, 2025 • 0 new comments

• Human in the loop

  #109 commented on Aug 29, 2025 • 0 new comments

• ModuleNotFoundError: openai.types.beta.realtime.conversation_item_input_audio_transcription_completed_event

  #1546 commented on Aug 29, 2025 • 0 new comments

• Calling agents as tools fails with "The tool returned an error" when using Gemini models

  #1565 commented on Aug 27, 2025 • 0 new comments

• Human-In-The-Loop Architecture should be implemented on top priority!

  #636 commented on Aug 27, 2025 • 0 new comments

• Error "Item ‘rs_ABCD’ of type ‘reasoning’ was provided without its required..." when using CodeInterpreter

  #985 commented on Aug 25, 2025 • 0 new comments

• Getting request tokens for the agent as a tool

  #1520 commented on Aug 25, 2025 • 0 new comments

• How do I handle image outputs from tools and feed it back to the LLM during an agent run?

  #1553 commented on Aug 25, 2025 • 0 new comments