microsoft/semantic-kernel · error · AgentExecutionException

{type(agent)} service failed to complete the request

Error message

{type(agent)} service failed to complete the request

What it means

Raised as an AgentExecutionException when agent.client.responses.create raises any Exception that is NOT a BadRequestError (i.e. not a 400). This is the catch-all for transport/auth/rate-limit/server errors from the OpenAI SDK during a non-streaming Responses create. The original exception is chained as the cause, so the real error is always inspectable via __cause__.

Source

Thrown at python/semantic_kernel/agents/open_ai/responses_agent_thread_actions.py:631

                instructions=merged_instructions or agent.instructions,
                previous_response_id=previous_response_id,
                store=store_output_enabled,
                tools=tools,  # type: ignore
                stream=stream,
                **response_options,
            )
        except BadRequestError as ex:
            if ex.code == "content_filter":
                raise ContentFilterAIException(
                    f"{type(agent)} encountered a content error",
                    ex,
                ) from ex
            raise AgentExecutionException(
                f"{type(agent)} failed to complete the request",
                ex,
            ) from ex
        except Exception as ex:
            raise AgentExecutionException(
                f"{type(agent)} service failed to complete the request",
                ex,
            ) from ex
        if response is None:
            raise AgentInvokeException("Response is None")
        return response

    @classmethod
    async def _poll_until_completed(
        cls: type[_T],
        agent: "OpenAIResponsesAgent",
        response: Response,
        polling_options: "RunPollingOptions",
    ):
        count = 0
        while response.status != "completed":
            await asyncio.sleep(polling_options.get_polling_interval(count).total_seconds())
            response = await agent.client.responses.retrieve(response.id)

View on GitHub (pinned to c028a0c7dc)

Solutions

1. Inspect the chained __cause__ exception to identify the exact HTTP status (401/429/5xx) and message.
2. For 401/403, verify the API key / Azure credentials and endpoint configuration.
3. For 429, implement backoff/retry (respect Retry-After) or reduce request frequency.
4. For connection errors, check network/proxy/SSL settings and endpoint reachability.
5. Update the openai SDK to a compatible version if an unexpected exception type is surfacing.

Example fix

# before
response = await agent.invoke(thread=thread)
# after - catch AgentExecutionException, inspect cause, retry on transient
from semantic_kernel.exceptions import AgentExecutionException
import openai
try:
    response = await agent.invoke(thread=thread)
except AgentExecutionException as ex:
    cause = ex.__cause__
    if isinstance(cause, openai.RateLimitError):
        await asyncio.sleep(backoff)
        response = await agent.invoke(thread=thread)
    else:
        raise

# Validate credentials/endpoint are set before invoking:
import os
assert os.environ.get("OPENAI_API_KEY"), "OPENAI_API_KEY not set"
assert agent.client.base_url, "client base_url not configured"

from semantic_kernel.exceptions import AgentExecutionException
import openai
async def invoke_with_retry(agent, thread, attempts=3):
    for i in range(attempts):
        try:
            return [m async for _, m in agent.invoke(thread=thread)]
        except AgentExecutionException as ex:
            cause = ex.__cause__
            if isinstance(cause, openai.RateLimitError) and i < attempts - 1:
                await asyncio.sleep(2 ** i)
                continue
            raise

When it happens

Trigger: responses.create throws e.g. openai.RateLimitError (429), openai.AuthenticationException (401), openai.APIConnectionError (network), openai.InternalServerError (5xx), or any other non-BadRequest SDK exception. Falls through the `except Exception` branch in _get_response.

Common situations: Invalid/expired API key (401); exceeding rate limits or quota (429); network outages/DNS failures to the API; OpenAI/Azure-side 5xx incidents; missing AZURE_OPENAI_ENDPOINT or wrong base_url; SDK version mismatch causing unexpected exception types; SSL/proxy misconfiguration.

Related errors

• Agent Failure - Run terminated: {run.Status} [{run.Id}]: {ru
• Agent Failure - Run not created for thread: ${threadId}
• Failed to retrieve messages for thread `{thread_id}`.
• Failed to search the collection.
• Astra DB not available. Status : {response}

AI-assisted analysis of microsoft/semantic-kernel@c028a0c7dc (2026-08-13). Data as JSON: /api/errors/9a9a46cce46d651d. Report an issue: GitHub.