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

Annotations already flag destructiveHint=true, openWorldHint=true, and non-idempotence, but do not say why; the description supplies the missing context that the call incurs real monetary cost and is gated on human approval. It stops short of explaining what side effects occur beyond billing or why the operation is non-idempotent, so it adds solid value without being complete.

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?

Two sentences, both load-bearing, with the cost warning and approval protocol placed after the core action so the agent reads what the tool does before how to sequence it. No filler or restatement of the name.

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 still indicates what comes back: an answer on an approved run and a cost on a preview call. It omits whether a preview returns anything else useful (such as an estimate identifier) and how to inspect or cancel the resulting run via siblings like get_run or cancel_run, which is a modest but real gap for a costly, non-idempotent operation.

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 coverage is only 33%, but the description compensates: it defines approved as a cost-preview/confirmation switch rather than a plain boolean, and explains that inputs is a per-{{variable}} map with get_prompt as the source of variable names. prompt_id is left implicit, which is a minor gap for an otherwise well-explicated three-parameter surface.

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?

"Run a deployed prompt once with values for its {{variables}} and return the answer" gives a specific verb (run), resource (a deployed prompt), execution mode (once), and the input convention ({{variables}}). It is clearly distinct from the sibling set's authoring tools (create_prompt, update_prompt, deploy_prompt) and from get_prompt, which it references for variable discovery.

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?

It states the exact two-step calling protocol: first call without approved to obtain the cost, then call with approved: true only after a human says yes. This is an explicit when-to-call-each-way rule, plus a pointer to get_prompt for enumerating variables, leaving nothing about the invocation sequence to inference.

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