The engineer is choosing between building a fully custom agent for the customs-compliance workflow versus using an Agent Bricks pattern. The workflow has highly bespoke, multi-step business logic specific to Meridian's internal escalation policy that no generic pattern anticipates. What does this imply about the right approach?
Select an answer to reveal the explanation.
Short Explanation
Managed patterns are great until your business logic gets weird enough that the pattern's assumptions stop matching reality - at that point, hand-building the chain yourself, with full control over each step, is the right call even though it's more work.
Full Explanation
A hand-built chain or agent is the more appropriate choice when the business logic is bespoke enough that no managed pattern's built-in assumptions match it well, because Agent Bricks patterns trade some flexibility for ease of setup, and that trade stops paying off once the workflow's specific escalation policy, sequencing, or decision logic falls outside what the pattern was designed to express. Always defaulting to Agent Bricks regardless of fit is wrong because 'least code' is not the only design criterion; forcing bespoke logic into a pattern that doesn't accommodate it produces a worse solution than building the exact chain the business needs, even at higher initial engineering cost. The claim that Agent Bricks patterns cannot call tools is factually wrong - patterns like Multi-Agent Supervisor and Knowledge Assistant are explicitly designed to incorporate tool use and orchestration, so 'no tool calling' is not a valid disqualifier. Abandoning the workload is an overreaction; a hand-built Mosaic AI Agent Framework chain is a fully viable and commonly used path for exactly this kind of custom logic, it is not a dead end. This item tests the reverse direction of the Agent Bricks decision: recognizing when NOT to reach for a managed pattern, not just when to reach for one.