You are defining how a web-search subagent handles failures so the coordinator can make good recovery decisions. Which two behaviors should the subagent implement? (Select two.)

Answer: A and B — Retry transient failures, such as a brief timeout, locally before reporting anything upward / When it cannot resolve a failure, return its type, what was attempted, partial results, and alternatives

Subagents should recover locally from transient failures and propagate only what they can't resolve, with structured context the coordinator can act on. (C) hides the context needed to choose a recovery. (D) terminates the whole workflow over one failure, discarding work that succeeded.