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

Annotations declare readOnlyHint=false and destructiveHint=true, and the description adds real context beyond them: searches are billed ('failed or empty searches leave no final charge, anything reserved is restored'), harvest is long-running ('tens of seconds to minutes'), semantic mode=deep is slower and costs more, and read auto-retries other fetch paths. It does not explain why the tool is flagged destructive/readOnly=false beyond the reservation/refund note, so a small gap remains.

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?

Front-loaded with the core sentence before the dispatch list, and each indented line carries one sub-tool with only the discriminating detail. Despite its length the text is dense with necessary routing information for 17 sub-tools, and the trailing warning is placed where it is read last but remembered.

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 needn't explain returns, and it does cover routing, cost/latency caveats and safety implications. The one real gap is that `params` is an untyped nested object, so the agent has no authoritative schema for each sub-tool's arguments and must infer them from inline hints.

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 `tool` enum and `params` object are nominally documented, but `params` is a free-form object ('The tool's params') with no per-sub-tool schema. The description compensates by naming concrete params inline (full, kind=patents, mode=deep, q/num in the example, and web's site/region/time-window/paging), adding meaning the schema lacks — though coverage of all 17 sub-tool param sets is incomplete.

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 opening sentence names the specific verb+resource (search and retrieve sources) and immediately discloses the dispatch mechanism ('17 tools, chosen with `tool`'), then enumerates every one with a distinct one-line purpose. An agent can distinguish crawl vs map vs harvest vs read without opening any other schema, and the tool is clearly separated from the account/billing/key siblings.

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?

Nearly every sub-tool line carries an explicit use/don't-use rule: 'use it only when crawl cannot get the data' (harvest), 'the opposite direction of media' (lens), 'something keyword search cannot do' (similar), 'use crosscheck to assert a specific fact'. The final warning line names the alternative tool for a concrete condition, which is exactly the when-to-use guidance an agent needs.

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