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

Annotations already declare readOnlyHint=true, destructiveHint=false, openWorldHint=false, but the description adds substantial behavioral context on top: only customers ≥10% of consolidated sales are disclosed (~40% of listed companies have a table, ~60% say '該当なし'), so an empty result does not mean no transaction, and the status enum (listed/none/noBlock/text/unparsed) is spelled out. It also discloses the amountYen vs amountRaw distinction for foreign-currency rows, the computed salesSharePct vs reported ratioPct, name masking for individuals, and the customerMatch=group caveat that group-level sales are not necessarily to the entity itself.

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-mode structure is front-loaded in a numbered list, followed by the disclosure-scope caveat and status table, with related tools and field-name conventions at the end. The trailing presentation guidance (metricLabels, Japanese-first field naming) is slightly beyond selection/invocation needs, but it is short and clearly separated; nothing in the core is redundant.

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?

There is no output schema, so the description carries the burden of describing return content, and it does: status values, amountYen/amountRaw, salesSharePct vs ratioPct, customerMatch, masked individual names, and metricLabels are all covered. For a tool whose main failure mode is misreading absent data as absence of a transaction, the disclosure-scope explanation makes it complete.

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% and the schema already documents all three parameters, but the description adds real meaning beyond it: it explains that `customer` accepts either a securities code (normalized to listed companies) or a partial company name (substring match that includes non-listed customers), which changes the result set fundamentally. It also clarifies that `limit` only governs the reverse-lookup direction, a constraint the schema does not state.

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?

States a specific verb and resource (returns '主要な顧客ごとの情報' from 有価証券報告書) and immediately splits into two precisely defined modes: code → that company's main customers, customer → reverse lookup of companies selling to that customer. The related-tools line (get_edinet_financial_data for financials, get_major_shareholders_history for shareholders) separates it cleanly from siblings without requiring the schemas to be opened.

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?

Explicitly says one of the two parameters must be chosen per call, explains what each mode returns, and gives concrete examples for the code-vs-name forms of `customer` (7203 vs アマゾン) along with the different matching behavior each triggers. It also states the exclusion conditions that would otherwise cause false conclusions — reverse lookup only covers exact-normalized names, so subsidiaries, non-listed customers, and spelling variants will not appear.

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