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

Annotations already declare non-read-only, non-idempotent, non-destructive, so the safety bar is partly met; the description adds real context beyond them: the run consumes GPU time on the sidecar, requires a claimed credential, and fails with 403 for unclaimed sandbox workspaces. It does not say what happens if a run is already in flight or how the result is identified.

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?

Three tight sentences, front-loaded with the action and parameters, then resource cost, then prerequisites. No sentence is filler; the sibling pointer and the 403 note each carry distinct information.

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, the description correctly directs the agent to get_aitune_profiling for results and covers auth and resource cost. The remaining gap is not stating what this call itself returns (e.g., a run identifier) or whether completion is signaled asynchronously.

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 100%, so the baseline is 3; the description adds that batch_sizes, measurement_steps, saturation_cutoff and include_warmup are optional overrides of defaults, which the schema alone does not convey. It does not add format or interaction details beyond that.

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 ('Start a profiling run for a model on the AITune sidecar') and explicitly distinguishes itself from the sibling get_aitune_profiling, which reads the finished report. An agent can separate the write-side profiling kickoff from the read-side report tool without opening either schema.

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 the follow-up path ('read the finished report with get_aitune_profiling') and hard preconditions: a claimed Last Price credential and a claimed workspace, with the sandbox-key 403 case spelled out as a de facto exclusion. It stops short of naming when profiling is unnecessary versus other AITune siblings such as list_aitune_backends or get_aitune_system_stats.

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