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

Annotations already cover readOnly/openWorld/idempotent/non-destructive, and the description adds substantial behavior beyond them: login is required, answers and contacts are masked by default (only hasContacts/contactKinds), PII embedded in free-text projections is deliberately scrubbed, ID numbers are never decrypted, limit is deliberately capped at 50 with rationale, and overview stats only appear on the first page. These are exactly the traps an agent would otherwise hit. No contradiction with annotations.

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?

Dense but scannable: bracketed section headers (【需要登录】【何时用】, 【组合链·批量处置】, 【口径/坑】) and a numbered ①-⑦ pitfall list earn their space for an 11-param tool. Minor deduction for the editorializing '碾压位' phrasing and the combo-chain paragraph being somewhat verbose relative to the tool's own scope.

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 an 11-param tool with no output schema, the description covers the return overview stats, per-row shape (question-key list, nickname, hasContacts/contactKinds), all masking defaults, and the pagination quirk. Remaining gaps are minor: no explicit sort order, no mention of a nextCursor field in the response, and no error/auth-failure behavior. The critical behavioral landmines are all disclosed.

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 itself is already descriptive (q's three-place search, fields' key-list default, reviewStatus enum meanings), so the baseline is 3. The description adds real operational value beyond it: q is best for project/company names, the concrete projection example fields=['project_intro'], the scrubbing caveat for projected answers, the rationale for the 50-row cap, and the overview-only-without-cursor pagination rule. Meaningful but incremental — the schema already carried most parameter semantics.

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 concrete trigger ('主办方问「报了多少人 / 今天新增几个 / 有哪些做 AI 的报了」时调它') and states the exact return shape (名单页 + 首页概览：总数/今日新增/待初审/渠道分布). It names the siblings it is not — get_signup_submission, issue_signup_export_link, bulk_review_signup_submissions — so an agent can route correctly without opening their schemas.

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?

Explicit when-to-use is front-loaded (organizer count questions), and exclusions are concrete: for contacting one person use get_signup_submission, for the full spreadsheet use issue_signup_export_link, and for batch disposition chain into bulk_review_signup_submissions with the preview=true → preview=false pattern. No inference is required.

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