research

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

Annotations already cover safety (readOnly, idempotent, non-destructive, open-world), and the description adds substantial behavior beyond them: latency per depth, what changes when a synthesis LLM is or isn't configured, the fallback instruction that the caller must write the answer from `brief`, and a recommendation to set up a provider key. That is exactly the kind of degradation-mode disclosure annotations cannot express.

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 purpose and differentiator, then behavior, parameters, and returns in clearly separated blocks. The parameter block is somewhat redundant with a fully-covered schema, but given the absence of an output schema the returns paragraph is not padding.

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 and a 10-parameter nested-input tool, the description carries its weight: it enumerates return fields (report, citations, sources, sub_queries, depth, total_time_ms, sampling_supported) and the `brief` sub-structure including the meaning of `gaps`. An agent knows what it will receive and how to consume it.

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 all ten parameters; the description's 'Key parameters' list largely restates them. It does add sub-query counts per depth (2/4/7) that the schema omits, but that is marginal value over the schema baseline.

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+resource ('Multi-step research on a complex question') and immediately describes the pipeline steps (decompose, parallel search, fetch, synthesize a cited report). It explicitly distinguishes itself from sibling tools by naming the `search` + `fetch` chaining pattern it replaces.

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?

'Beats chaining `search` + `fetch` manually for multi-source synthesis' names the alternative and the condition that favors this tool, and the depth tiers give timing/scope guidance for choosing invocation cost. No explicit when-not case (e.g., single-fact lookups should use `search`), so it stops short of full routing guidance.

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