Fixed permission issue that's causing backwards compatibility test run to fail & made logic to run prepare mode tests more robust

[adityabharadwaj198]

• What kind of change does this PR introduce? (Bug fix, feature, docs update, ...)
  Bug fix

• What is the current behavior? (You can also link to an open issue here)
  Compatibility tests are not running in marqo-ai/marqo repo due to providing overly permissive permission (i.e write-all) (https://github.com/marqo-ai/marqo/actions/runs/11951180519). These tests were earlier run on a forked repo and worked fine: (https://github.com/adityabharadwaj198/marqo/actions/runs/11927467398). While investigating the issue I found that some logic of how we run our prepare method could be improved, so I've fixed that as well.

• What is the new behavior (if this is a feature change)?
  Marqo compatibility tests will run in marqo-ai/marqo repo w/o failures

1. (https://github.com/marqo-ai/marqo/actions/runs/11954641816/)

• Does this PR introduce a breaking change? (What changes might users need to make in their application due to this PR?)

• No

• Have unit tests been run against this PR? (Has there also been any additional testing?)

• Added sample test run link:

• 1. (https://github.com/marqo-ai/marqo/actions/runs/11954641816/)

• Related Python client changes (link commit/PR here)

• None

• Related documentation changes (link commit/PR here)

• None

• Other information:

• Please check if the PR fulfills these requirements

• The commit message follows our guidelines
• Tests for the changes have been added (for bug fixes/features)
• Docs have been added / updated (for bug fixes / features)