Add default include of reasoning.encrypted_content for store=False to prevent reasoning errors

[ihower]

Issue

When using store=False without including reasoning.encrypted_content, any function calls (including handoff) that generate reasoning items will result in an error because the server does not persist reasoning items. This makes the Responses API difficult to use in practice.

I believe it would be better if reasoning.encrypted_content were included by default in this case, since in almost all scenarios the agent needs it when working with reasoning models and store=False.

Related issue: openai/openai-agents-python#1668

Reproducible Example

from agents import Agent, Runner, function_tool, ModelSettings

@function_tool
def get_weather(city: str) -> str:
    return "hot"

agent = Agent(
    name="Test store false",
    instructions="You are a helpful assistant.",
    model="gpt-5-mini",
    tools=[get_weather],
    model_settings=ModelSettings(store=False)
)

result = Runner.run_sync(agent, input="What's the weather in Hsinchu?")

print(result.final_output)

This results in:

Error getting response: Error code: 404 - {'error': {'message': "Item with id 'rs_xxx' not found. Items are not persisted when `store` is set to false. Try again with `store` set to true, or remove this item from your input.", 'type': 'invalid_request_error', 'param': 'input', 'code': None}}. 

Fix

• When model_settings.response_include is None and model_settings.store is False, automatically add "reasoning.encrypted_content" to the include set.
• This change makes it easier for developers using gpt-5, since they no longer need to worry about missing reasoning data when store=False.

[seratch]

@codex review this

[seratch]

Thanks for the suggestion but I don't think this is robust enough

[seratch]

non-reasoning models such as gpt-4.1 do not require this (actually it does not work), so always having this does not work. why don't you manually add this option to include?

[ihower]

You are right, I forgot to consider that GPT-4.1 would throw errors, and this solution will cause breaking changes, which is not good. However, I do hope there’s a better way to help developers realize that when Store=False, they need to include reasoning.encrypted_content, otherwise GitHub issues will keep showing up.

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. More of your lovely PRs please.

About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you open a pull request for review, mark a draft as ready, or comment "@codex review". If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex fix this CI failure" or "@codex address that feedback".