I don’t want overpromising; I need realistic guidance on what tasks are suitable for an agent and what isn’t.

Provide clear, realistic guidance on which tasks are appropriate to delegate to an agent and which are not. This helps set accurate expectations and reduces wasted time, rework, and risk from misusing automation.

This feature gives practical, grounded guidance on when an agent is a good fit and when it isn’t. It helps users evaluate tasks based on factors like clarity of requirements, tolerance for errors, need for human judgment, and potential impact if something goes wrong. By clarifying limitations and best-fit scenarios, it reduces overpromising and prevents users from delegating work that is unlikely to succeed. The guidance can be used early in planning to decide whether to automate, partially automate, or keep a task fully manual. It also supports better scoping by encouraging users to define success criteria, inputs, and constraints before handing work to an agent. Common suitable use cases include repetitive, well-bounded tasks with clear instructions and verifiable outputs. Common unsuitable use cases include high-stakes decisions, ambiguous objectives, or work requiring deep domain authority without reliable validation. The end result is higher trust, better outcomes, and more efficient use of agents by matching tasks to appropriate automation levels.