Skip to content

Errors and outcomes

BaseHarbor uses typed outcomes so humans and agents can distinguish different failure classes.

Important semantic outcomes include:

  • unsupported;
  • unverifiable;
  • denied / not authorized;
  • conflict;
  • foreign ownership;
  • degraded;
  • failed verification.

Errors should be actionable and secret-safe.

Machine interfaces return structured error envelopes. Human CLI rendering may add context, but it must not change the underlying semantic result.