Best way to identify input error for multi_match?

Elastic Stack Elasticsearch
1

We’re passing user input into a multi_match with minimal escaping. If the user searches ”something (unbalanced quotes), ES responds with a 400:

{
  root_cause: [
    {
      type: 'token_mgr_error',
      reason: 'token_mgr_error: Lexical error at line 1, column 10.  Encountered: <EOF> after prefix "\\"something" (in lexical state 2)'
    }
  ],
  type: 'search_phase_execution_exception',
  reason: 'all shards failed',
  phase: 'query',
  grouped: true,
  failed_shards: [ ... ],
  caused_by: {
    type: 'token_mgr_error',
    reason: 'token_mgr_error: Lexical error at line 1, column 10.  Encountered: <EOF> after prefix "\\"something" (in lexical state 2)'
  }

I can’t find documentation on exceptions. Can I reliably assume search_phase_execution_exception indicates user input error or should I check for token_mgr_error - or something else? I’d like to distinguish user-input issues from deeper issues.

Thanks!