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

Annotations mark this readOnlyHint=false with no destructive hint, and the description reconciles that by disclosing the side effect: with run_id it records to the run report and can fail the run. It also discloses a hard prerequisite (Labs Marketing and offer emails must be enabled) and the resulting error behavior, plus a known measurement limitation (dot trick). It stops short of describing the full response shape or what happens on partial failures.

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 content is dense and mostly earns its place (reuse semantics, Labs gate, dot-trick exclusion), but the first sentence is a fragment that forces re-reading, and the Labs paragraph is long for a conditional setup note. Front-loading is decent—reuse detection appears early—but the prose could be tightened considerably into shorter declarative sentences.

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 no output schema, the description carries return-shape duty and does name the key fields (reused, same_code, run_id effect) and the reuse condition. It also covers the setup prerequisite, an error mode, and a limitation. Missing: what a 'sender' scope contains and whether results are per-run accumulative, which an agent might want before calling.

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 already 100%, so baseline would be 3; the description earns above that by explaining run_id's consequence (a failing run is recorded against that run) and by defining the output flags reused and same_code, which no structured field covers. The address-variant behavior is also restated meaningfully ('acts on the plain address'). Test is left entirely to the schema.

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 conveys that this tool reports, per sender, the coupons received at an address and which address variant each went to, and names its sibling alternative (get_address_variants) for signing up with variants. The opening clause is grammatically broken ('For each sender, its coupons at this address and which version...') with no verb, so the exact operation is inferred rather than stated. Still, an agent can distinguish it from check_email/check_delivery/count_emails by the offer-reuse focus.

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?

Explicitly routes the agent to get_address_variants for creating variants, states the Labs prerequisite that must be enabled before the call will succeed, and rules out the Gmail dot trick as untestable here. It stops short of saying when NOT to call it (e.g. when no one-time offer is involved) or of sequencing itself relative to start_run/finish_run.

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