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

Annotations already declare readOnlyHint=true, idempotentHint=true, and destructiveHint=false, covering safety. The description adds context beyond annotations by specifying that it reveals account-specific capabilities and constraints (supported task types, spend caps, allowed area, KRW balance, pricing), which are behavioral traits of the tool. The 'Read-only' at the end reinforces the annotations without contradiction. Since annotations cover the safety profile, the description's added context about what the tool reveals (e.g., account limitations) contributes value. Score 4 because it adds richer context than just repeating the annotations; it explains what the tool returns and its purpose in a pre-task context.

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?

The description is a single sentence that is front-loaded with the use-case trigger ('Use before a local-human or physical-world task'), followed by the specific resource contents, ending with 'Read-only.' Every word earns its place: it includes the trigger, the list of what is returned, and the safety hint. No filler or redundancy. Efficient and well-structured.

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?

Given that there is no output schema, the description carries the burden of explaining what the tool returns, which it does effectively by listing the items. For a zero-parameter, read-only capability tool, it covers the essential information: when to use, what it tells you, and that it's read-only. It could mention that it returns results for the current account (implicit) or cost implications, but given the simplicity, 4 is appropriate. It doesn't explain response format, but the tool is arguably a simple capability query. A score of 4 reflects it's complete for its purpose; only missing minor details like exact response structure.

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?

The input schema has zero parameters, so there are no parameters to document. The description doesn't need to explain parameters. A baseline of 4 is given for zero-parameter tools because the schema provides no parameter information. The description does not attempt to describe parameters (there are none), which is correct. It focuses on the output content. Score 4 aligns with the rule: '0 params = baseline 4' – it fully covers what an agent needs to know about parameters (there are none).

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?

The description uses a specific verb ('Use before a local-human or physical-world task') and names the resource ('Errand') and clearly enumerates what it returns: supported on-site task types, spend caps, allowed area, KRW balance, and pricing. It distinguishes itself from siblings by describing a capabilities/capability discovery tool, whereas siblings like errand_dispatch or errand_quote perform actions or estimations. The phrase 'to see what Errand can do for this account' frames it as a discovery capability tool, differentiating it from action-oriented siblings.

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?

The description explicitly states when to use: 'Use before a local-human or physical-world task' – giving a clear contextual trigger. It lists the specific items it returns, which helps the agent determine if it needs this information. While it does not explicitly name sibling alternatives, the collective sibling list makes it evident that this is a pre-flight check, not an action. The guidance is clear and actionable, though it does not mention when NOT to use it, but the 'before a task' framing implies it's a prerequisite check, not a substitute for action tools.

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