Cannot delete a AI Foundry Project Index (Standard Agent Setup)

[skillbuilderzone]

• Package Name: azure.ai.projects
• Package Version:
• Operating System:
• Python Version: 3.13

Describe the bug
The Index operation is unable to delete the 'Project Indexes' in a standard Agent setup.

To Reproduce
Steps to reproduce the behavior:
A Standard Agent Setup in Azure AI Foundry Project.
Note : This is NOT a hub based project but the Standalone project with Standard BYO setup.
-Create an Agent in a standard setup (Not the hub based project but the standalone project)
-Add knowledge using an Azure AI Search Index (This is an external azure ai search service)

• Project Index gets created .
• Now to delete any Project Index we used the sdk as from the AI Foundry Portal we could not find the option.

Expected behavior
The project index (alias/pointer/wrapper/) gets deleted from the Project and cannot be further found in the list of existing indexes
Screenshots
The below does not delete the indexes while it can list and show the versions fine.

Additional context
Add any other context about the problem here.

[github-actions]

Thanks for the feedback! We are routing this to the appropriate team for follow-up. cc @dargilco @glharper @howieleung @jhakulin @nick863 @trangevi.

[dargilco]

@skillbuilderzone thank you for opening this issue. I'm Darren from the SDK team that owns azure-ai-projects and azure-ai-agents client libraries. To help with the investigation, I'd like to ask you to run a simple script that gets the dataset, then tries to delete it, while console logs turned are turned on. Please do the following:

Copy this code into a *.py file:

import os
from azure.identity import DefaultAzureCredential
from azure.ai.projects import AIProjectClient

endpoint = os.environ["PROJECT_ENDPOINT"]
index_name = os.environ["INDEX_NAME"]
index_version = os.environ["INDEX_VERSION"]

with DefaultAzureCredential(exclude_interactive_browser_credential=False) as credential:

    with AIProjectClient(endpoint=endpoint, credential=credential) as project_client:

        print(f"Get Index `{index_name}` version `{index_version}`:")
        index = project_client.indexes.get(name=index_name, version=index_version)
        print(index)

        print(f"Delete Index `{index_name}` version `{index_version}`:")
        project_client.indexes.delete(name=index_name, version=index_version)

Set the environment variables to define your project endpoint and existing dataset name and version (the one you were not able to delete):
set PROJECT_ENDPOINT=
set INDEX_NAME=
set INDEX_VERSION=

Set this environment variable to turn on unredacted logging:
set ENABLE_AZURE_AI_PROJECTS_CONSOLE_LOGGING=true

Run your code.

Copy the console output as a text file.

e-mail the result to dcohen@microsoft.com (my e-mail). Don't share the resulting log in this GitHub issue... e-mail it to me instead.

Thanks,

Darren

[skillbuilderzone]

@skillbuilderzone thank you for opening this issue. I'm Darren from the SDK team that owns azure-ai-projects and azure-ai-agents client libraries. To help with the investigation, I'd like to ask you to run a simple script that gets the dataset, then tries to delete it, while console logs turned are turned on. Please do the following:

Copy this code into a *.py file:

import os
from azure.identity import DefaultAzureCredential
from azure.ai.projects import AIProjectClient

endpoint = os.environ["PROJECT_ENDPOINT"]
index_name = os.environ["INDEX_NAME"]
index_version = os.environ["INDEX_VERSION"]

with DefaultAzureCredential(exclude_interactive_browser_credential=False) as credential:

with AIProjectClient(endpoint=endpoint, credential=credential) as project_client:

    print(f"Get Index `{index_name}` version `{index_version}`:")
    index = project_client.indexes.get(name=index_name, version=index_version)
    print(index)

    print(f"Delete Index `{index_name}` version `{index_version}`:")
    project_client.indexes.delete(name=index_name, version=index_version)

Set the environment variables to define your project endpoint and existing dataset name and version (the one you were not able to delete): set PROJECT_ENDPOINT= set INDEX_NAME= set INDEX_VERSION=

Set this environment variable to turn on unredacted logging: set ENABLE_AZURE_AI_PROJECTS_CONSOLE_LOGGING=true

Run your code.

Copy the console output as a text file.

e-mail the result to dcohen@microsoft.com (my e-mail). Don't share the resulting log in this GitHub issue... e-mail it to me instead.

Thanks,

Darren

Thanks Darren. have sent the logs while redacting some info.