Inspect a humanoid rig and its animation donors

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

The description adds a useful safety claim—'Reads only caller-selected local files and downloads nothing'—which goes beyond the read-only/idempotent annotations. But it is not fully transparent: the `target` object variant allows 'acquire a pinned sample into the development cache,' which implies a cache write and possibly fetching, undercutting the blanket 'Reads only' statement. Because the annotations themselves declare read-only behavior, this is a descriptive overreach rather than an annotation contradiction.

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 a single well-organized sentence with a front-loaded action/resource followed by a compact list of returned data, plus one short sentence for the access/security behavior. It contains no fluff and every clause earns its place.

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 two-parameter read-only inspection tool with an output schema, the description covers the main inputs and safety behavior compactly. However, it misses the `sourceId` pinned-sample input form, the cache-write implication, and any explicit relationship to sibling tools, leaving an agent with an incomplete mental model of when and how to invoke it. The output schema covers return values, so no deduction there, but the input-mode gap prevents a higher score.

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 `target` and `libraries`; the description mostly repeats those facts. It also narrows `target` to 'a local humanoid GLB' and omits the pinned-sample/sourceId acquisition option, which could mislead an agent about valid inputs. This is not a meaningful semantic addition beyond the schema.

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 names the action ('Inspect'), the exact resource ('local humanoid GLB and optional local UAL ZIP/GLB libraries'), and enumerates the outputs (meshes, joints, suggested bone mappings, clips, attributes, extensions, measured bounds, attachment candidates). This makes it clearly distinct in purpose from sibling mutation/preview/import tools such as asset_auto_rig and asset_retarget_animations.

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 implies the tool should be used when an agent needs detailed structural/animation-donor information before rigging or retargeting, and its read-only privacy claim clarifies safe contexts. However, it never explicitly states when to choose this over asset_auto_rig, asset_preview_animation, or asset_retarget_animations, and it gives no exclusions or prerequisites. Usage guidance is only implied, not stated.

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