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

The annotations declare idempotentHint=false, but the description says 'Idempotent transactional messages' and states that replaying the same key returns the original receipt and is not re-executed. This is a direct annotation contradiction and per rubric scores 1. The description does add useful context about channel fallback and honest failure, but the idempotency contradiction is severe.

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 organized into a short definition, a concrete example, WHEN TO USE, WHEN NOT TO USE, COST, LATENCY, and EXECUTION. This is well structured and front-loaded. The 'do not use for marketing' point appears twice and the text contains a mojibake artifact, but overall the length is justified by useful operational context.

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 idempotency, channel fallback, honest failure behavior, cost, latency, and directs the agent to get_outcome for the result, which is valuable for a tool with no output schema. However, it does not define the result or error shape, and the parameter example contradicts the schema, leaving an agent without a reliable request template. For a tool with 5 parameters and nested objects, this is not fully complete.

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?

Schema description coverage is only 40%, so the description needed to compensate. The example call uses a recipient object with id_type and id_value, while the input schema defines recipient as an object with phone_or_email (and optional name). Additionally, the booking_confirmation data example uses business_name while the schema's data description uses smb_name. These mismatches are actively misleading. The idempotency_key and confirmation_type context is helpful, but the wrong examples undermine correctness.

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 identifies the resource as transactional messages (OTP, booking confirmation, payment receipt, cancellation notice) and the example query shows exactly how to invoke it. The WHEN NOT TO USE section separates it from marketing and conversational messaging, making it distinguishable from send_message. The only small weakness is that the opening is a noun phrase rather than an explicit verb, but the tool name and example make the action unambiguous.

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?

Explicit WHEN TO USE (messages that MUST be delivered reliably) and WHEN NOT TO USE (marketing or conversational) are present. The exclusions are specific enough for an agent to avoid misrouting, though it does not name the sibling tool that should be used instead. The guidance is clear but not maximally explicit about alternatives.

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