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

Beyond the annotations (readOnly/idempotent/non-destructive), it discloses a lot the annotations cannot: retry.afterSeconds semantics with null ending polling, the meaning of a 'finalizing' state, that only completed results are deliverable, and the distinct return shapes per tool type (subtitle transcription yields a documentId/revision for get_subtitle_document; rendering yields a saved video). This is exactly the extra behavioral context a long-running-operation poller needs.

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 core action, then layers polling, state, and output semantics in short sentences with no filler. The telegraphic clauses ('Follow retry.afterSeconds; null ends automatic polling.') are dense yet each earns its place, though the density borders on requiring a second read.

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-param read-only poller with no output schema, the description covers what an agent must know: when to poll, how to interpret retry.afterSeconds and finalizing, and what the returned artifacts are per tool. Missing pieces such as not-found/error behavior and pagination are minor here.

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% and both params are required, so the description carries the burden. It indirectly hints at toolId values by describing what 'subtitle transcription' and 'rendering' produce, but never explains operationId's meaning or maps the enum values to behavior. It partially compensates for the coverage gap but leaves it incomplete.

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 ('Read the existing operation') and scopes it ('scoped to this account and OAuth client'), which separates it from the generation-status tools. It also clarifies the negative ('Does not submit generation'), so an agent won't confuse recovery with submission. It stops short of a fully crisp one-line identity but is clearly not tautological.

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 concrete when-to-use instructions: 'Recover after disconnects before proposing another attempt,' plus a polling policy ('Follow retry.afterSeconds; null ends automatic polling'). The 'does not submit generation' line implies when not to use it. It never explicitly names an alternative sibling to switch to, so it is not a full 5.

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