submit_inspection

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

No annotations are provided, so the description carries the full behavioral burden. It does contribute one genuinely useful trait — no forward passes are executed, so there is no inference cost — and 'queue' hints at asynchronous execution. Beyond that it is silent on whether the call blocks, what resource limits or permissions apply, what happens when weight_reads is null, and whether anything is mutated or persisted.

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 short sentence with the action verb front-loaded and no filler. It is efficient, though the terse phrasing comes at the cost of precision rather than being trimmed of redundancy.

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?

An output schema exists, so return values need not be explained, but the tool pairs a richly nested request schema with zero annotation coverage and a one-line description. An agent cannot determine from the definition what the returned job handle looks like, how to retrieve results, or how to construct a valid weight_reads entry, all of which the description should at least sketch.

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% across four parameters, including a nested WeightReadSpec with roughly eleven undocumented fields (target, source, invocation, coordinates, slices, max_elements, token_positions). The only hint the description offers is 'bounded parameter reads', which loosely gestures at max_elements/slices but gives no meaning to source, invocation, offset, or the interaction between indices and coordinates. For a schema this dense, the description adds far too little.

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 verb 'queue' plus the phrase 'without forward passes' makes clear this is an asynchronous job submission that does no inference, which is a useful distinction from the read-side sibling get_inspection. However, 'model catalog' does not map cleanly onto any schema field (the schema only has a 'model' ModelSpec and 'weight_reads'), so an agent must guess that it means 'load/describe a model'. The purpose is inferable but stated in internal jargon rather than in terms an agent can match to the parameters.

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?

There is no when-to-use guidance, no statement of prerequisites, and no mention of alternatives such as submit_experiment or get_inspection. The description never says that this returns a queued job whose results must later be fetched, leaving the agent to discover the follow-up call on its own.

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