🪲 MT.1024 Not parsing recommendations correctly for tests

[CrookedJ]

Describe the bug

The Entra recommendations tests are not executed due to the result of the graph request being parsed incorrectly. This results in a failed test that is malformed.

To Reproduce

Steps to reproduce the behavior:

1. Execute Invoke-Maester -Path .\Maester\Entra\ -Tag Recommendation
2. Results show one malformed (ID-less) test has failed

Module Version

2.0.0

Environment Data

Name                           Value
----                           -----
PSVersion                      7.5.4
PSEdition                      Core
GitCommitId                    7.5.4
OS                             Microsoft Windows 10.0.19045
Platform                       Win32NT
PSCompatibleVersions           {1.0, 2.0, 3.0, 4.0…}
PSRemotingProtocolVersion      2.3
SerializationVersion           1.1.0.1
WSManStackVersion              3.0

Screenshots

Additional context

The graph request in .\Maester\Entra\Test-EntraRecommendations.Tests.ps1 to retrieve the recommendations is $EntraRecommendations = Invoke-MtGraphRequest -DisableCache -ApiVersion beta -RelativeUri 'directory/recommendations?$expand=impactedResources' -OutputType Hashtable

This results in $EntraRecommendations containing a hashtable like

Name                           Value
----                           -----
value                          {0064a2ca-615a-4bca-9351-d2b35242dc12_aadGraphDeprecationServicePrincipal, 0064a2ca-615a-
@odata.context                 https://graph.microsoft.com/beta/$metadata#directory/recommendations(impactedResources())

Yet the test attempts to loop through that hashtable and reference non-existent properties

<#..snip..#> -ForEach $EntraRecommendations {
$RecommendationId = $_.id`

Modifying it to -ForEach $EntraRecommendations.value { so it loops through them correctly does work, but the resulting IDs seem wrong. I'm unsure of what the correct IDs are.

[merill]

Good catch. Thanks for raising this @CrookedJ

[stefandoh]

@CrookedJ, @merill, imho the fix is already suggested bei @CrookedJ.
I checked the commits and seems the commit 72eb353 refactored the test with purpose to not use sequence numbers any longer.

Earlier, those test gave an MT.1024.1, MT.1024.2, etc and falling back to MT.1024.becRetirement if there was no key assignment. I am agreeing with that commit to just keep the subtests ids.