Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Beyond the annotations (readOnlyHint=false, destructiveHint=false), the description discloses the concrete side effect: 'This action creates an auditable reconciliation record.' It also explains canonicality, idempotency expectations, and instructs the agent to trust the returned actual_tool_output and context_engineering instead of adding a state-recovery call.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is front-loaded with the core purpose and call condition, then organized into labeled Canonicality, Idempotency, and Confirmation boundary sections. It is dense and mostly non-redundant, though phrases like 'plus every prerequisite stated above' add little and the wording is somewhat jargon-heavy.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The description covers prerequisites, provider evidence, side effects, idempotency, confirmation requirements, and the two schema branches—strong coverage for a tool with no output schema. It stops short of describing the returned payload or success/failure indicators, but it does tell the agent to trust the returned actual_tool_output.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The description does not document each field individually, but it maps the two schema branches to the two allowed provider-confirmation states: exact onboardingDataId/signup link versus no active application/retry safe. This gives meaningful semantics to the resolution consts and the handoffUrl/providerApplicationId pair that the bare schema properties do not convey.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with 'Owner/operator recovery for one uncertain Mercury onboarding attempt,' giving a specific verb, resource, and scope. This clearly separates it from status-read siblings like get_bank_onboarding_status and domain-different siblings like reconcile_payment_route, while also stating that the action 'creates an auditable reconciliation record.'

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description provides explicit call conditions: 'Call only after Mercury directly confirms either the exact onboardingDataId and mercury.com signup link, or that no active application remains and a corrected retry is safe.' It also gives a hard prerequisite, a confirmation boundary requiring fresh explicit user confirmation, and a warning never to invent provider evidence—leaving little ambiguity about when invocation is permitted.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.