Artificial intelligence agent task scoping turns a request into work an agent can execute and verify. A well-scoped task identifies the desired outcome, allowed changes, relevant context, constraints, risks and evidence required for completion.
Ambiguous decisions should remain collaborative until people establish intent and tradeoffs. Delegating uncertainty as if it were a clear implementation task can produce technically polished work that solves the wrong problem.
