拼多多联系客服（读/受控写）

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

Annotations cover the safety profile (readOnlyHint=false, destructiveHint=false), so the bar is lower, yet the description still adds substantive behavior: write mode is off unless the full guard set is present, and web/customer-service text must be treated as data rather than authorization — valuable prompt-injection guidance the annotations cannot express. It omits failure behavior and what the acceptance gate actually checks. The "唯一请求 ID" wording sits in mild tension with idempotentHint=false, though a dedup key does not strictly make a call idempotent, so this is not a contradiction.

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, mode semantics front-loaded before the security caveat, with no filler. Every clause carries information. The undefined "验收闸门" costs a point because it is the one phrase an agent cannot act on.

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 6-parameter tool with an output schema and full annotations, the description supplies the mode model, the write guard set, and the injection-resistance rule — the pieces structured fields cannot carry. What remains missing is the meaning of the acceptance gate and what happens when prerequisites are unmet. That is a small but real gap rather than a blocking one.

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 83%, so the schema already explains most fields and the baseline is 3. The description earns above baseline by stating the mode-conditional requirement — that shop_id, message, confirm and request_id matter only in send mode — a dependency the schema cannot express, since it lists only goods_id as required. It still does not explain the format or provenance of shop_id verification.

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 names the two operating modes with specific verbs: "read 读取已核对店铺的会话" and a gated "send" path, so an agent can tell what the tool does in each mode. The stated scope (会话 of an already-verified shop) is more precise than the title alone. It stops short of naming what distinguishes this from siblings like pdd_read_item or pdd_search, so it is clear but not sharply differentiated.

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?

"send 默认关闭" plus the enumerated prerequisites (shop_id, message, confirm, request_id) tells the agent the default path is read-only and what must be gathered before a write. That is real when/how guidance, not just a purpose statement. It never names an alternative tool or an explicit when-not-to-use case, and the referenced "验收闸门" gate is never defined, so the guidance is strong but incomplete.

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