microsoft/semantic-kernel · error · ValueError

The prompt execution settings must not have a response forma

Error message

The prompt execution settings must not have a response format set.

What it means

StandardMagenticManager parses the model's progress/facts/plan ledgers via structured output, so it sets response_format on the prompt execution settings itself during construction. If the settings object you pass already carries a non-null response_format, the manager's own structured-output contract would be overwritten or made ambiguous, so the constructor rejects it. This is a configuration guard, not a runtime failure.

Source

Thrown at python/semantic_kernel/agents/orchestration/magentic.py:257

                - task_ledger_facts_prompt: The prompt to use for the task ledger facts.
                - task_ledger_plan_prompt: The prompt to use for the task ledger plan.
                - task_ledger_full_prompt: The prompt to use for the full task ledger.
                - task_ledger_facts_update_prompt: The prompt to use for the task ledger facts update.
                - task_ledger_plan_update_prompt: The prompt to use for the task ledger plan update.
                - progress_ledger_prompt: The prompt to use for the progress ledger.
                - final_answer_prompt: The prompt to use for the final answer.
        """
        # Bast effort to make sure the service supports structured output. Even if the service supports
        # structured output, the model may not support it, in which case there is no good way to check.
        if prompt_execution_settings is None:
            prompt_execution_settings = chat_completion_service.instantiate_prompt_execution_settings()
            if not hasattr(prompt_execution_settings, "response_format"):
                raise ValueError("The service must support structured output.")
        else:
            if not hasattr(prompt_execution_settings, "response_format"):
                raise ValueError("The service must support structured output.")
            if getattr(prompt_execution_settings, "response_format", None) is not None:
                raise ValueError("The prompt execution settings must not have a response format set.")

        super().__init__(
            chat_completion_service=chat_completion_service,
            prompt_execution_settings=prompt_execution_settings,
            **kwargs,
        )

    @override
    async def plan(self, magentic_context: MagenticContext) -> ChatMessageContent:
        """Plan the task.

        Args:
            magentic_context (MagenticContext): The context for the Magentic manager.

        Returns:
            ChatMessageContent: The task ledger.
        """
        # 1. Gather the facts

View on GitHub (pinned to c028a0c7dc)

Solutions

1. Pass prompt_execution_settings=None and let the manager instantiate fresh settings from the service.
2. Before passing settings, clear the field: settings.response_format = None.
3. Use a chat completion service/connector that exposes the response_format attribute so the manager can manage structured output itself.

Example fix

// before
settings = AzureChatPromptExecutionSettings(response_format=JsonSchemaResponseFormat(my_schema))
manager = StandardMagenticManager(service, prompt_execution_settings=settings)  # raises

// after
settings = AzureChatPromptExecutionSettings()
manager = StandardMagenticManager(service, prompt_execution_settings=settings)

# Before constructing StandardMagenticManager
settings = chat_completion_service.instantiate_prompt_execution_settings()
if not hasattr(settings, "response_format"):
    raise ValueError("Service must support structured output (response_format attr).")
if getattr(settings, "response_format", None) is not None:
    settings.response_format = None  # let the manager own structured output
manager = StandardMagenticManager(chat_completion_service, prompt_execution_settings=settings)

def is_clean_for_magentic(settings) -> bool:
    return hasattr(settings, "response_format") and getattr(settings, "response_format", None) is None

try:
    manager = StandardMagenticManager(service, prompt_execution_settings=settings)
except ValueError as e:
    if "response format" in str(e):
        settings.response_format = None
        manager = StandardMagenticManager(service, prompt_execution_settings=settings)
    else:
        raise

When it happens

Trigger: Constructing `StandardMagenticManager(chat_completion_service, prompt_execution_settings=settings)` where `getattr(settings, 'response_format', None) is not None`. Also triggered when settings lack the `response_format` attribute entirely (different message, same method).

Common situations: Reusing a PromptExecutionSettings instance that was configured for a separate structured-output call (json_schema/auto). Copying settings from another function that set response_format. Using a connector whose settings default response_format to a non-null value.

Related errors

• All members must have a description.
• The service must support structured output.
• The members list cannot be empty.
• Configuration not found, please setup the notebooks first us
• AZURE_OPENAI_ENDPOINT is not set.

AI-assisted analysis of microsoft/semantic-kernel@c028a0c7dc (2026-08-13). Data as JSON: /api/errors/8ea3cf82946ded8f. Report an issue: GitHub.