Doco Search V2

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

The annotations already declare readOnlyHint=true, idempotentHint=true, and destructiveHint=false, so the description's job is to go beyond safety traits — and it does substantially. It discloses the response composition, the cursor-based exhaustive traversal semantics, and critically the negative-evidence rule: when projection.complete=false, results are incomplete and '不能据此断言知识不存在' (cannot assert knowledge absence). This directly prevents a classic agent failure mode of treating a search miss as proof of non-existence. No contradiction with the annotations; it refines the open-world nuance without conflicting.

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 dense sentences with no filler: the first front-loads the purpose and enumerates the return fields, the second conveys the exhaustive-mode traversal capability and the completeness caveat. Every clause earns its place, and the most decision-relevant warning (incomplete results cannot prove absence) is placed at the end where it reads as a caution.

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?

An output schema exists, so the description is not obligated to fully document return values, yet it still summarizes the key return categories and adds the completeness semantics that no schema could express. For a 6-parameter tool with pagination, an enum mode, and locale/KB scoping, this covers the essentials. The one genuine gap is sibling routing: with doco_search present in the same tool list, the absence of any statement about which search variant to prefer leaves a meaningful completeness hole.

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 description coverage is 100%, so the schema carries the per-parameter documentation burden and the baseline is 3. The description adds genuine value on the mode/cursor interplay — that exhaustive mode combined with cursor allows complete traversal — which the schema's terse 'mode parameter' and 'opaque pagination cursor' text does not convey. It does not, however, add meaning for q, limit, locale, or knowledge_base_id beyond what the schema already provides.

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 a specific verb and resource — '全文搜索' (full-text search) — qualified by a distinctive feature, '带查询级完整性证明' (with query-level completeness proof), and enumerates the return content (catalog path, title path, context, score explanation, watermarks). However, it never differentiates itself from the near-identically named sibling doco_search; the reader must infer why two search tools exist rather than being told.

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 context is implied rather than explicit: 'exhaustive 模式可用 cursor 完整遍历' tells the agent that exhaustive mode plus cursor enables full traversal, and the projection.complete=false caveat indicates when a negative result is not trustworthy. But no alternative tools are named, and the obvious sibling doco_search is not addressed with any 'use this when / use that when' guidance, leaving the selection between the two search tools to inference.

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