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

Despite the annotations only indicating non-read-only and non-destructive, the description goes far beyond. It discloses login requirements, the de-duplication rule ('resolves to the newest revision'), the never-collapse behavior for different 고지, the meaning of requires_account_link=False, and two distinct error/partial-result shapes. This is substantial behavioral context that an agent cannot infer from annotations or schema.

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 dense paragraph that packs purpose, auth, dedup rules, account-link behavior, and error semantics. While every sentence adds value, it lacks visual structure and would be easier for an agent to parse if broken into bullets or short paragraphs. It is not wasteful but is longer than necessary for a tool definition.

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 a complex composite tool with 7 parameters and no output schema, the description covers the operational flow, preconditions, merging logic, and partial failure modes. The only notable gap is that it never explicitly describes the successful output envelope structure, even though it references 'computed', 'computed_error', and 'legs'. Still, the coverage is strong and materially compensates for the missing output schema.

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 already provides 100% coverage with descriptions for all 7 parameters, so the baseline is 3. The description adds meaningful domain context: it explains 고지 and 작성회차, how row keys disambiguate matches, and that goji_round relates to the notice round. This deepens understanding of how parameters like nhis_row_key and nps_row_key should be used, going beyond the schema's surface-level descriptions.

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 opens with 'Composite premium summary v1: searches NHIS and NPS premium notices, then passes each selected row to the corresponding premium-summary Job.' This is a specific verb (searches, passes, generates) and resource (premium summary for NHIS/NPS). It clearly distinguishes itself from sibling summary tools like employment_insurance_summary_v1 because it explicitly names its inputs (health insurance and pension).

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 states a clear precondition: 'Requires prior session login for 건강보험·국민연금.' It also explains when it is applicable (for producing a composite premium summary) and how it handles edge cases. However, it does not explicitly mention alternatives or exclusions when not to use it, but the purpose is clear enough to infer usage from the tool name and sibling context.

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