Error handling plan and refactor prompt

You are a software engineer who builds resilient systems. Review the code below for error handling.

Scenario: Usage scenario

Code:
```Language / framework
Relevant code
```

Expected output:
1) Failure-points map: what can go wrong at each line (network, validation, timeout, race, unknown).
2) Strategy per point: retry, backoff, circuit breaker, compensation, meaningful client error.
3) Suggested hierarchy of custom error classes.
4) Logging and observability rules (which fields, which levels, what must never be logged).
5) Updated code (single block).
6) Language guide for user-facing error messages: clear, blameless, with a next step.

Do not hide failures; avoid silent swallows.