queryData

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

Annotations declare readOnlyHint, idempotentHint, and destructiveHint=false, and the description aligns and reinforces these with 'deterministic, read-only'. It adds beyond annotations by disclosing truncation behavior ('Large results are truncated for context; the full result stays retrievable'), the authentication requirement ('free community login'), the seed requirement for simulations, and the offline capability. No contradictions; the description enhances the safety and behavior profile.

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 concise for the tool's complexity—about 90 words—and front-loads the core purpose and kind routing before diving into details. Each sentence earns its place: kind distinctions, bindings advice, seed requirement, truncation handling, and auth. The only minor inefficiency is that the truncation detail could be seen as part of output behavior, but it is necessary context. It is dense but well-organized.

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?

The description is complete despite the tool's high complexity (two distinct schemas, QA vs computation). It explains both usage modes, the necessity of listSources for field discovery, the truncation-and-getResult recovery path, auth requirements, and the seed requirement for simulations. All critical invocation aspects are covered, and the absence of an output schema is mitigated by mentioning getResult and result_id. Nothing an agent needs to call it correctly is missing.

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% and includes descriptions for metric, source, group_by, module, function, bindings, seed, and filter fields. The description adds semantic context beyond the schema: it explains the conceptual distinction between query and computation kinds, advises using bindings instead of pasting data, and mentions the seed requirement explicitly. It doesn't enumerate every parameter name but adds strategic guidance on how to use them, which is valuable above the schema's individual descriptions.

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 specific verb ('Execute'), a specific resource ('one deterministic, read-only SQAI request'), and distinguishes two usage kinds (query vs computation) with concrete examples. It explicitly names sibling tool listSources for discovery, preventing confusion with getResult and explainQuery. The purpose is unambiguous and actionable.

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 gives explicit when-to-use guidance for each kind ('Use kind 'query' for tabular questions (aggregations, grouping, filtering)', 'Use kind 'computation' for math (statistics, financial, vector/matrix, simulation)'), states the alternative for large data (bindings vs pasting), and mentions the getResult tool for retrieving truncated results. It also covers prerequisites (listSources first, login requirements) and identifies when not to use it (never paste large data).

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