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

Annotations already declare readOnlyHint=false, destructiveHint=false, and openWorldHint=false, but the description adds material context the annotations cannot convey: each call consumes one token, the call creates an XBOSS task and persists its processing result, and it is explicitly not a pure read-only operation regardless of whether the source institution record is merely queried. That is a genuinely useful side-effect disclosure that prevents an agent from treating this as a free lookup. It stops short of 5 because the mixed English/Korean text and the boilerplate-like closing sentence about institutional data changes make the operational impact less crisp than it could be.

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 purpose is front-loaded, which is good, but the body is a long run of idempotency boilerplate that reads as templated across many tools and includes an odd phrase ('anonymous taste calls') that weakens confidence. The abrupt switch from English to Korean mid-paragraph, plus the generic closing line about per-tool confirmation procedures, adds length without proportional clarity.

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 only two optional parameters, full schema coverage, and no output schema, the definition covers the call mechanics and side effects adequately. What is missing is any indication of what user information comes back and an explicit authentication prerequisite, both of which an agent would need since there is no output schema to fall back on. Adequate, but with clear gaps for a read-style tool whose payload is undescribed.

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 100%, so baseline is 3, but the description adds real meaning beyond the schema for request_idempotency_key, whose schema entry only says 'see the tool description' and therefore delegates its semantics here. The generate-save-reuse-new-key lifecycle and the warning not to confuse it with a provider-specific idempotency_key are exactly the kind of guidance the schema cannot carry. account_link_id is handled entirely by its own schema description, which is appropriate.

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 first sentence gives a specific verb and resource: retrieve logged-in user info from Hometax. That is unambiguous about what the tool returns conceptually. It does not, however, distinguish itself from sibling session tools such as hometax__session__login or the various session__status tools, so an agent must infer that this is the post-login identity read.

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?

Usage guidance exists but is scoped entirely to the idempotency key: reuse only to recover a lost response, use a new key for a new execution, and the key is unavailable to anonymous calls. There is no guidance on when to call this tool versus sibling session or account-link tools, nor an explicit statement that a login must precede it. The implied 'call after authentication' context is left to inference.

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