A flow can encounter a failed record update, and the team needs a controlled response rather than an unexplained failure. What should be configured?

1. Appropriate fault paths and error handling
2. Permanent removal of all validation and access controls
3. An assumption that activated flows cannot fail
4. A longer flow label as the only remedy