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

Annotations already declare readOnlyHint=true, idempotentHint=true, destructiveHint=false and openWorldHint=false, so the safety profile is covered. The description adds real behavioral context by revealing that several kinds are scoped to a parent entity (a page's diagrams, a React app's files/code), which implies those calls need parent context in params. It says nothing about result shape, pagination, or result limits, which for a 10-kind dispatcher is a remaining gap.

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?

A single front-loaded sentence states the purpose, then the kind list carries the detail with compact parentheticals; the closing note about params is one short clause. Dense but every token earns its place, and nothing is repeated from the schema's own descriptions.

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?

For a polymorphic 10-kind tool with an untyped params bag and no output schema, the description covers the 'what kind' decision but leaves the 'which filters' decision to a guide() round-trip the agent must infer. Safety behavior is covered by annotations, but result shape and required parent identifiers for dependent kinds are not spelled out.

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%, but the params object is opaque (additionalProperties, no sub-fields documented), so the description is doing meaningful work by glossing the discriminator values: designs means keyword-searching screens and React apps, react_files means listing an app's files, react_code means searching an app's code. That distinction directly affects choosing the right kind. The filter payload itself is entirely deferred to guide(), so the params semantics remain partially unexplained.

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 first clause names a specific verb and domain ('Find UX Pilot work'), and the kind list enumerates exactly which resource types are reachable, including how each is scoped ('diagrams (a page's)', 'react_code (search a React app's code)'). What's missing is any contrast with siblings like read, catalog, or get_job, so an agent cannot tell from the description alone why it would pick find over those.

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?

Usage is only implied: the enumeration of kinds suggests this is the search/listing entry point across entity types, and 'Filters go in params' hints at where narrowing happens. There is no explicit when-to-use vs. when-not, and no mention of read/catalog as alternatives for single-resource retrieval.

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