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

Annotations already declare the tool as read-only and non-destructive, so the description does not need to repeat that safety profile. It adds significant behavioral detail beyond annotations: it describes the exact return shape (columns, column_types, rows), edge cases (all-zero rows omitted, hidden_zero_rows), how ids align on summary vs detail reports, and the meaning of filters and no_data. This goes well beyond the minimal expectations and provides deep transparency, though it could clarify whether pagination or limits exist, which is not mentioned.

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 description is a single, dense paragraph that front-loads the primary purpose and then efficiently details the return format. Each sentence contributes substantive information about behavior or output structure. It is not overly long and there is minimal wasted phrasing, though it could be broken into smaller sentences for easier scanning. Still, it is well within acceptable length and avoids redundancy.

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?

Given the tool's moderate complexity (a report with nested structures), the description covers essential aspects: the action, the input parameters (via schema), the output format, edge cases, and error-like conditions. The absence of an output schema makes this explanation critical, and it does provide it. However, it lacks explicit mention of pagination limits or potential throttling, and doesn't clarify how errors (e.g., invalid date) are surfaced, but for this context it is quite complete.

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?

The input schema already provides a comprehensive description of the params object, covering date formats, macros, accounting_method, summarize_column_by, and filters, with 100% coverage. The description does not add new parameter-specific details beyond what's already in the schema. It references 'filters' but without explaining them further, so it adds minimal value. Given the schema's thoroughness, a baseline score of 3 is appropriate.

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 states a clear action ('Generate the QuickBooks class sales report') and a specific resource (class sales). It distinguishes itself from related report tools (like get_customer_sales, get_item_sales) by starting with 'class sales,' making the purpose unambiguous. The detailed return format description also reinforces what the tool produces, setting it apart from simply listing data.

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?

The description mentions that filters are applied and that ignored parameters are absent from the output, which implies how to use it (e.g., if a filter isn't applied, it won't appear). However, it doesn't explicitly state when to use this tool over alternatives (e.g., versus get_customer_sales, get_item_sales) or provide exclusions. The sibling list shows many similar report tools, but no guidance on selecting among them is given, leaving the agent to infer based on the 'class sales' focus.

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