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

Beyond readOnlyHint/openWorldHint, it discloses the comparison baseline (period vs. the period before), the default window edge ('periods end yesterday except today'), and the goal-aware judgement rule so lead/awareness campaigns are not mislabeled waste. It also states that each finding carries the supporting sentence, numbers, and a daily savings figure. These are exactly the behavioral details that prevent an agent from misreporting results.

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 lead ('the four questions people ask most') front-loads the value well and nearly every clause carries differentiating content. However, the body is one sprawling run-on sentence with deeply nested parenthetical lists, which makes it harder to scan than a structured breakdown into 'what's included' and 'when to call it' would be.

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 the return-value burden and does so by specifying the four finding categories, their per-item supporting data, and the savings estimate. It covers goal-dependent judgement, which is the main failure mode for this tool. Minor gaps remain: behavior with no data in the period, and that relative periods anchor to the account timezone.

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 schema already documents since/until/period and the last_7d default, making 3 the baseline. The description adds only the boundary semantics ('periods end yesterday except today'), which clarifies how the default and relative periods resolve but says nothing about custom-range behavior or how since/until interact with period.

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 enumerates exactly what the report contains: spend by campaign/ad set with shares, winners judged against account cost, ads ranked by cost per result with link CTR and hook rate, and waste/fix items. That is a specific verb-plus-resource statement an agent can match to a request. It stops short of 5 because it never names or contrasts a sibling (e.g. get_ad_insights, get_account_health), so differentiation rests on the agent inferring from content alone.

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?

Explicit invocation triggers are given in quoted user phrasings: 'how are my ads doing', 'what should I stop', 'what is working', 'which creative is best', plus the pre-budget-recommendation case. That is strong positive routing. It lacks any when-not guidance or named alternative for narrower questions (e.g. single-campaign or pixel/health questions), so it falls short of 5.

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