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

Annotations already declare readOnly, idempotent and non-destructive, so the safety profile is covered; the description still adds useful behavior: nothing is persisted, the call is repeatable, and unreachability is broken into named reason categories (no address, suppressed, existing relationship, in another campaign, cooling down). Missing only that these counts consume no credits/allowance, which sibling tools like getContactAllowance imply matters here.

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 sentences, front-loaded with the core count-plus-reasons behavior and the no-op guarantee. The main waste is the duplicated guarantee — 'Saves nothing.' followed by 'It saves nothing, so it can be repeated.' — which restates the same fact instead of adding a new one.

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?

For a read-only counting tool with a deep nested schema and no output schema, the definition covers the essentials: what is counted, what the result explains, and that it is side-effect free. It would be stronger if it stated the response shape (a count plus a per-reason breakdown) or noted that filters combine as an AND, since the agent has no output schema to fall back on.

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 must carry the load. It maps the audience sources well (requests, ICPs, companies, contacts, pipeline) and documents the interaction of include_pipeline with available_again, but says nothing about the large filters object (states, signals, countries, industries, departments, seniorities, company_sizes, sub_industries), new_only, recontact_lead_id, or destination_requirement. Roughly half the input surface remains undocumented.

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 ('counts how many contacts a selection of requests, ICPs, companies, contacts or the pipeline could reach now') and adds the distinctive second output, why the rest can't be reached. That verb+resource pairing separates it from mutation siblings like createAudience or saveLeadEngineAudience. It stops short of naming the closest read sibling (previewAudience, getAudiencePreflight) explicitly.

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 gives one concrete scenario — finding past contacts worth a new campaign by sending include_pipeline with available_again — plus the reassurance that repeating is safe. There is no general statement of when to use this versus previewAudience or getAudiencePreflight, and no explicit exclusions, so the agent must infer the boundary from intent.

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