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

Annotations only cover safety/mutability flags; the description adds real behavioral context the agent could not infer — that the chat client fills in the temporary download URL, that convert.express downloads from it, and that the tool is asynchronous (returns job_id, then poll get_job). It omits the non-idempotency implication (repeated calls spawn new jobs, consistent with idempotentHint=false) and any cost/auth expectations, so it earns a 4 rather than 5.

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?

Four short sentences, zero filler. The primary action and its environment constraint come first, the alternative routing second, and the follow-up step last — perfectly front-loaded for an agent scanning for actionability.

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 an 8-parameter, nested-object, async tool with no output schema, the description supplies the missing lifecycle piece (job_id then get_job) that the schema cannot express. It leaves unstated whether jobs incur cost or must be paid via pay_job, and does not note that repeated calls create separate jobs, which are small but real gaps given the sibling tools available.

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%, including the nested file object, model enum guidance pointing at get_service_info, and the robot_key explanation, so the schema carries the parameter burden. The description adds nothing parameter-specific beyond the download_url mechanics already documented in the schema, making baseline 3 correct.

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 transcribing an audio or video file") plus the scope constraint that the file must be attached in a ChatGPT chat. The sibling set includes transcribe_url, start_transcription, and create_upload, and the description makes clear this one is the attached-file path without needing to open any 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?

Articulates the selection condition (file attached in a ChatGPT chat) and explicitly routes the agent elsewhere for the main alternative case: "Other chat apps cannot fill in that link: use create_upload for files attached there." It also tells the agent what to do next after the call (poll get_job).

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