consult

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

Goes well past the annotations: it explains that my_view is deliberately withheld from the remote model so answers stay independent and is echoed back for comparison, that lenses are answered independently in parallel, and that results are UNVERIFIED output from a weaker model to be weighed below your own analysis. This is exactly the behavioral context the readOnly/openWorld hints cannot convey.

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-loads the purpose and trigger in two short lines, then uses a compact per-parameter list where every line adds distinct information. The closing warning about unverified suggestions is a third short sentence that earns its place by changing how the agent should weigh results.

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 an output schema present, the description correctly omits return-value details and instead covers cost model, independence semantics, and result trust level. The one omission is the 'thinking' parameter (high/max), which an agent could not infer from the description alone.

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 description coverage is 0%, so the description carries the full burden and does it: question, context (including what to include and what was already ruled out), lenses with a concrete example and a hard cap of 4, my_view with its unusual non-transmission semantics, and project_dir with a precise conditional ('only if Deputy picked the wrong folder'). Only the 'thinking' enum parameter is undocumented, which is a minor gap against five well-specified parameters.

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 and resource — brainstorm with a cheaper model to surface ideas, risks and alternatives — and adds a scope note ('not on every task') that separates it from exhaustive siblings like review or investigate. It stops short of naming a sibling tool to contrast against, so the differentiation is inferred rather than explicit.

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?

Gives a clear trigger ('when you are unsure between approaches or stuck') and an explicit exclusion ('not on every task'), which most definitions omit. It never points at the sibling tools (review, investigate, delegate) as the alternative for cases outside this trigger, leaving the routing decision partly to inference.

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