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

Annotations (readOnlyHint=false, openWorldHint=true, idempotentHint=false, destructiveHint=false) carry little information, and this description carries the full burden with rich disclosures: login required, per-record model billing with a no-loop-retry warning, cross-anchor dedup keeping highest score plus all reasons, one full-site search per leadId with 429 rate-limit consequence, and null-leadId candidate handling. The scare-quoted '这个「读」' framing aligns with readOnlyHint=false rather than contradicting it — it explains that despite feeling like a read, the operation has side effects (cost).

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 every sentence earns its place. The critical cost warning is front-loaded with a ⚠ marker, and the four labeled sections 【需要登录】【两种用法】【组合链】【口径】 make a long description highly scannable. No filler or repetition of schema content.

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?

Given a complex, billed, rate-limited tool with no output schema, the description covers the essentials: cost, rate limits, dedup semantics, null-leadId handling, usage modes, and the surrounding workflow. It references output semantics (leadId, bUserId, score, reasons) but stops short of specifying the exact return structure, which is a minor gap for a tool of this complexity.

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% with decent parameter descriptions, so the baseline is 3. The description adds meaning beyond the schema: the two usage-mode recipes directly prescribe how to set leadIds/perLead for different goals, and the '每条 leadId 都是一次全站检索，撞 429 就是扫太猛了' note connects parameter count to cost and rate-limit behavior. This elevates it above baseline, though the schema already covers basic meaning so it does not reach 5.

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 states a specific verb (scan/find candidates) and resource (broker leads), detailing the two merged retrieval paths (private-pool complementary scoring + full-site search), cross-anchor deduplication, and score/reason retention rules. It clearly differentiates from siblings via the combination chain list_broker_leads → scan_broker_matches → create_broker_match, so an agent can tell it apart from create_broker_match, list_broker_leads, and introduce_broker_match 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?

The 【两种用法】 section gives two explicit parameter recipes for distinct intents (pool inventory = many leadIds/small perLead vs. hard-match grinding = single leadId/perLead 20) and flags that the perLead=20 mode is impossible in App/web. The 【组合链】 section names the exact workflow with sibling tools and instructs how to phrase the reason field in create_broker_match, and the 'don't loop retry' rule sets an explicit exclusion. Nothing 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.