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

Annotations already declare a non-read-only, non-destructive, non-idempotent, closed-world operation; the description adds real context on top by disclosing that a durable draft is persisted, a draft_id is returned, nothing is sent, and the call consumes the org's AI token budget. It doesn't address retry behavior (e.g. whether a repeated call creates a second draft) or failure modes, which keeps it from a 5.

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?

Three tight sentences with the action and its inputs first, then the persistence/never-sends guarantee, then cost — no filler and nothing buried.

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?

With an output schema present, return details needn't be restated, and the description still flags draft_id and the safety/cost profile, so an agent can call it correctly. The remaining gap is edge behavior — what happens with an invalid conversation, missing AI configuration, or budget exhaustion.

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 0%, so the description has to carry the burden; it explains that 'instruction' is an optional free-text steer for the draft, which is genuinely useful. However, it says nothing about the conversation_id format constraint (numeric pattern) or the 500-character cap on instruction, leaving part of the gap for the schema alone.

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 names a specific verb and resource (draft a reply for a conversation) and qualifies the mechanism — org AI tone, presets, knowledge context — which separates it from read-only siblings like list_messages or get_conversation. It also states the output artifact (draft_id) and rules out sending, so an agent knows exactly what it gets.

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?

It makes the use case clear: produce a draft for human review in the inbox, with an optional instruction to steer it. There are no explicit when-not conditions or named alternative tools, but the 'NEVER sends' clause effectively defines the boundary between this tool and any send action.

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