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

Annotations only say it is a non-read-only, non-destructive, open-world tool; the description adds materially more, namely that the default mode is non-impacting (observation) while TARGETING is the reach-changing variant, and that it 'runs directly' without a confirmation step. The Performance Max limitation is another real behavioral constraint. It does not mention reversibility of an applied targeting audience or any rate/limit behavior.

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?

One tight, front-loaded passage: action, mode semantics, execution behavior, then the exclusion. Every clause carries decision-relevant information and there is no boilerplate.

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 7-parameter mutation tool with no output schema, the description covers purpose, mode risk, and the major unsupported case, which is most of what an agent needs to call it safely. Minor gaps remain around which ID parameter to supply and what a success response contains, though dry_run is self-documented in the schema.

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 57%, and the description compensates on the most consequential parameter by explaining what OBSERVATION vs TARGETING actually do, which the bare enum does not. It also tells the agent where the audience IDs come from. It adds nothing about user_list_ids vs user_interest_ids beyond the schema's own hints, nor about campaign_id/ad_group_id selection.

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?

States a specific verb and resource ('Add audience segments ... to a campaign or ad group') plus the scope of both placement targets. It also names the upstream sibling google_ads_find_audiences as the source of IDs, so an agent can distinguish creating/locating audiences from attaching them.

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?

Clearly frames the two modes and their effect — observation changes nothing about reach, targeting narrows reach — which is the main decision an agent faces. It also excludes an unsupported case (Performance Max) and points to the UI alternative. It stops short of saying whether campaign_id or ad_group_id is required, or whether they are exclusive, so the routing guidance is not fully closed.

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