Live tests: ResilienceManagement test-resources deployment hangs and is cancelled by worker timeout (~60min), reproducible across runs

## Summary

The `Live tests - toolsAzure.Mcp.Tools.ResilienceManagement` CI job consistently fails because the **"🚀 Deploy test resources"** step never finishes provisioning test infrastructure — it is still polling ARM ("Checking deployment status...") when the ~60 minute CI worker timeout kills the job. This has now been observed in **two consecutive runs**, so it is a persistent/reproducible failure, not a one-off flake.

## Evidence

Example: [Run #20260910.4](https://dev.azure.com/azure-sdk/internal/_build/results?buildId=6812627&view=logs&j=3062cd78-3dd8-5cfa-d581-ea116ba27da8&t=39f71255-8c7a-511e-e12f-47ec4d48350d&l=66)

| Run | Deploy started | Job cancelled (worker timeout) | Elapsed |
|---|---|---|---|
| build 6800475 (run1) | 2026-09-08T21:21:52Z | 2026-09-08T22:27:03Z | ~65 min |
| build 6812627 (run2) | 2026-09-10T23:22:05Z | 2026-09-11T00:29:39Z | ~67 min |

Both runs show the exact same pattern:
```
##[error]The Operation will be canceled. The next steps may not contain expected logs.
##[error]The operation was canceled.
##[section]Finishing: 🚀 Deploy test resources
...
##[error]Step cancelled due to worker timeout
##[error]Job cancelled due to worker timeout.
```

## Suspected root cause

`tools/Azure.Mcp.Tools.ResilienceManagement/tests/test-resources.bicep` deploys, among faster/standard resources (Storage account, Managed Disk, Automation account):

```bicep
resource usagePlan 'Microsoft.AzureResilienceManagement/usagePlans@2026-04-01-preview' = {
name: usagePlanName
location: 'global'
properties: {
planType: 'Standard'
}
}
```

## Environment

- Location: `westus`
- Template: `tools/Azure.Mcp.Tools.ResilienceManagement/tests/test-resources.bicep`
- Post-deploy script: `tools/Azure.Mcp.Tools.ResilienceManagement/tests/test-resources-post.ps1`

Start with tools/Azure.Mcp.Tools.ResilienceManagement/tests/test-resources.bicep and tools/Azure.Mcp.Tools.ResilienceManagement/tests/test-resources-post.ps1, then review the linked CI run's deployment logs. Reproduce the Live tests - toolsAzure.Mcp.Tools.ResilienceManagement job and determine why ARM remains in "Checking deployment status...". Done means test-resource deployment completes within the worker limit across repeated runs.

Written by the indexing model from the issue text.