Freigeben über

Azure AI Foundry Agents no longer visible in portal and Agent Application endpoint returns HTTP 500

Admin Aebischer Etienne STAWI 0 Zuverlässigkeitspunkte
2026-05-03T10:13:08.3733333+00:00

Hello Microsoft Support / Community

since yesterday evening, our Azure AI Foundry agents are no longer visible in the Azure AI Foundry portal. Before that, the same agents and endpoints worked successfully from an Azure Logic App.

We are using Azure AI Foundry Agents with a published Agent Application endpoint, which is called from Azure Logic Apps via Managed Identity.

Issue:

  1. Agents are no longer visible in the Azure AI Foundry portal.
  2. The existing published Agent Application endpoint now returns HTTP 500.
  3. A direct API test against the agent endpoint also returns HTTP 500.
  4. The issue also occurs with a minimal test payload, so it does not seem to be related to the Logic App body, JSON format, or document content.
  5. The same setup worked before and stopped working without changes on our side.

Error returned in Azure Logic Apps:

{

"error": {

"code": "server_error",

"message": "An error occurred while processing your request. You can retry your request, or contact us if the error persists. Please include the request ID e7db032c1b17ae2dd0d11b0226f2b505 in your message.",

"type": "error",

"details": [],

"additionalInfo": {

  "remove PII"

}
  }

}

Request ID: 


The endpoint pattern we are using is:

https://<foundry-resource>.services.ai.azure.com/api/projects/<project-name>/applications/<agent-application-name>/protocols/openai/responses?api-version=2025-11-15-preview

Test payload:

{

  "input": "Test"

}

Authentication:

- Azure Logic Apps HTTP action

- Managed Identity

- Audience: [https://ai.azure.com]()

- Content-Type: application/json

Permissions checked

- The user has Owner and Azure AI User permissions.

- The Logic App system-assigned managed identity has Azure AI User permissions on the Foundry project.

- Therefore, this does not look like an IAM/RBAC issue.

- If this were an authentication or authorization issue, I would expect HTTP 401 or 403, but the service returns HTTP 500.

Impact:

Our Logic App workflow can no longer invoke the published Azure AI Foundry Agent Application. This blocks our automated document processing workflow.

Could you please investigate whether there is an issue with Azure AI Foundry Agents, Agent Applications, or the agent runtime backend? The provided request ID is from the failed Logic Apps call.

Thank you.
Foundry Tools
Foundry Tools

Früher bekannt als Azure-KI-Dienste oder Azure Cognitive Services; eine einheitliche Sammlung vordefinierter KI-Funktionen innerhalb der Microsoft Foundry-Plattform


3 Antworten

Sortieren nach: Am hilfreichsten
  1. Anshika Varshney 11,310 Zuverlässigkeitspunkte Externe Microsoft-Mitarbeiter Moderator
    2026-05-07T01:15:59.7366667+00:00

    Hi Admin Aebischer Etienne STAWI,

    Thank you for sharing the detailed findings. This helps to clearly understand what is happening.

    From your observation, the agent itself is still available because the GET API is working and returning the agent configuration. This means the agent is not deleted and the configuration is stored correctly.

    However, the issue starts when you try to invoke the agent runtime. Both runtime endpoints are returning HTTP 404, which means the execution endpoint is not available. In simple terms, the system can see the agent definition, but it is not able to run it.

    This situation usually indicates a runtime availability or routing issue, not a data issue. So the agent exists, but the system is not able to map or expose the runtime endpoint for it.

    From what you described, it also confirms that the problem is likely related to the project state. Since you mentioned that the old project seems broken and you had to create a new one, it suggests that the old project may be in an inconsistent state. In such cases, it is possible that agent metadata is still present, but the runtime layer is not properly connected.

    Another important point is that you are using the preview feature with the header: Foundry-Features: AgentEndpoints=V1Preview

    Since these endpoints are still in preview, there can be situations where the agent configuration API works, but the runtime endpoint is not fully provisioned or exposed in that project.

    The fact that both tested routes return 404 also suggests that the runtime endpoint is not registered for that agent in the project, rather than a request formatting issue.

    In your case, creating a new project and being able to create agents again is a strong indicator that the issue is isolated to the old project. Continuing with the new project is a good approach.

    You can also make sure in the new project that:

    The agent is created and tested in the same project context

    The model deployment is properly linked to the agent

    The region and resource are aligned for project, model, and agent

    These are required for the runtime to work correctly.

    So overall, based on your findings:

    Agent configuration exists and is retrievable

    Runtime endpoint is not available

    Cause is likely project-level inconsistency or runtime mapping issue

    Using a new project resolves the issue

    This is a very useful observation, and your testing clearly shows the difference between agent configuration and runtime availability.

    Hope this helps. Do let me know if you have any further queries.

    Thankyou!

    War diese Antwort hilfreich?

    Eine Person fand diese Antwort hilfreich.

  2. Admin Aebischer Etienne STAWI 0 Zuverlässigkeitspunkte
    2026-05-11T18:26:11.14+00:00

    Hi sorry for my late response. The agents are visible again but I still had to change the way I am using the agent in Logic Apps. I had to adapt to a new URI and do not use the preview API anymore.

    Since then it has been working fine again.

    War diese Antwort hilfreich?

    0 Kommentare Keine Kommentare

  3. Admin Aebischer Etienne STAWI 0 Zuverlässigkeitspunkte
    2026-05-03T11:21:17.4066667+00:00

    Additional finding: I created a new project in AI Foundry so I could create Agents again, because the old project seems somehow broken.

    The agent itself can be retrieved successfully via:

    GET /api/projects/<project>/agents/<agent-name>?api-version=v1

    This returns the agent configuration including the system prompt.

    However, invoking the agent runtime fails with HTTP 404 for both tested routes:

    POST /api/projects/<project>/agents/<agent-name>/protocols/openai/v1/responses?api-version=v1

    and

    POST /api/projects/<project>/agent/protocols/openai/v1/responses?api-version=v1

    Both calls use:

    • Bearer token for https://ai.azure.com
    • Content-Type: application/json
    • Foundry-Features: AgentEndpoints=V1Preview
    • Body: {"input":"Test"}

    This suggests that the agent management API is reachable, but the published Responses runtime endpoint is not available or not routed correctly.

    War diese Antwort hilfreich?

    0 Kommentare Keine Kommentare

Ihre Antwort

Antworten können von Fragestellenden als „Angenommen“ und von Moderierenden als „Empfohlen“ gekennzeichnet werden, wodurch Benutzende wissen, dass diese Antwort das Problem des Fragestellenden gelöst hat.