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

Annotations only assert destructive/idempotent/no-read; the description adds far more: exactly what is destroyed (files, packages, background tasks, memory), irreversibility, billing semantics (stops, final partial period billed, per-second with 60 s minimum for ephemeral, per-started-hour for persistent), and the token requirement.

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-loaded with the destructive consequence, then the recovery caveat (download_file first), then billing, then auth requirement. Every clause carries operational weight; nothing is padding.

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?

No output schema exists, and none is needed for a delete. Between the description and annotations, an agent knows the safety profile, the scope of data loss, the billing effect, and the auth prerequisite — everything needed to invoke it correctly.

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?

There is a single parameter and schema coverage is 100%, with the schema itself documenting the id pattern and example. The description's only parameter-adjacent line ('Requires: token') concerns auth, not sandbox_id, so it adds no semantic detail 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?

Opens with a specific verb+resource+mode: 'Delete a sandbox IMMEDIATELY and IRREVERSIBLY (running or paused)'. That phrasing separates it from the read-only siblings get_sandbox/list_sandboxes and from kill_task, which targets a task rather than the sandbox itself.

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 a clear pre-condition: 'download_file anything you need first', which tells the agent the ordering constraint before invoking. It does not explicitly name an alternative path (e.g. pause or cancel_service) or state when deletion is unnecessary, so it falls short of full when/when-not guidance.

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