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

Annotations only mark it as a non-readonly, non-idempotent, non-destructive create, so the description carries most of the weight and delivers: it discloses the reply-only limitation on hosted addresses, the 30-day forwarding window for replaced mailboxes, and that omitted local_part is generated. The prompt-injection warning is genuinely useful agent-facing behavior guidance rather than a restatement of annotations.

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?

Purpose and the two domain modes are front-loaded, followed by usage and then the safety note, so an agent gets the essentials immediately. The untrusted-content sentence is longer than strictly necessary but is high-value guidance for a tool that ingests inbound mail.

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 values need no explanation, and all nine parameters are documented in the schema. What the description leaves out is the tenant relationship and any precondition that a tenant or verified domain must already exist, which matters for a create tool with zero required parameters.

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 coverage is 100%, so the baseline is 3; the description earns an extra point by explaining consequences the schema does not, namely that omitting domain yields a hosted reply-only address and that local_part is auto-generated when absent. It does not explain the tenant parameter's role or whether it is required, which is the main semantic gap.

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 first sentence gives a specific verb and resource plus the scope that matters: a real mailbox the agent can both read and send from. It then splits the behavior into two concrete modes (hosted domain = reply-only, verified domain = can initiate mail), which distinguishes it from siblings like create_domain, create_sender, and create_tenant.

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?

"Use it to sign up for a service or to take email from people" gives a clear use case, and it explicitly routes the forward_from case to forward_mailbox. It does not state when-not to use it (e.g., needing a sender identity instead, or requiring a tenant first), so it stops short of full when/when-not coverage.

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