[AZINTS-4596] Add user consent flow before logging unexpected script errors to Datadog

[mshvartsberg]

Motivation

When the Azure integration quickstart script encounters an unexpected error, we want to log the full traceback to Datadog to aid debugging but only with explicit user consent since the traceback may contain PII.

Changes

• script_status.py: Added FAILING_AWAITING_USER_DECISION status, USER_DECISION_TIMEOUT_SECONDS constant, and _poll_for_user_decision() method. On unexpected errors, the script now posts failing_awaiting_user_decision, then polls the workflow status endpoint for up to 60s waiting for a user_error_decision step with status finished. If the user approves error log forwarding, the full traceback is sent in the message field via a failed status; if they decline or the poll times out, no traceback is sent (message is None).
  • Also updated USER_ACTIONABLE_ERROR to be defined as "user_actionable_error" instead of "USER_ACTIONABLE_ERROR" to match the status definition in web-ui and dd-source
• test_script_status.py: Replaced the single test_unexpected_error test with three cases covering consent approved, consent declined, and poll timeout.

Testing

• Unit tests cover all three consent flow outcomes (approved, declined, timeout)
• Both end to end flows for configure log forwarding and app registration still work (on a regular flow with no errors).
• TO-DO: End-to-end test: trigger an unexpected error and verify the consent modal appears and each path (forward logs / file support ticket / click X / timeout) results in the correct final status