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

Annotations only supply idempotentHint/destructiveHint; the description adds the login prerequisite, partial-update semantics ('未传不变' - unspecified fields stay unchanged), a caution not to touch SMS/email unprompted, and the state-transition rules for icebreak=false and snoozeUntil=null/true-clear. That is substantive behavior beyond structured data, and nothing conflicts with destructiveHint=false since these are setting toggles rather than deletions.

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?

Very compact with no filler and the login requirement plus patch rule front-loaded, so every clause earns its place. However, the telegraphic style (missing spaces, abbreviated clauses, mixed Chinese/English field names) raises parsing cost and would benefit from clearer clause separation.

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?

With 12 all-optional params, full schema coverage, and no output schema, the description covers the risky parts: auth, partial updates, protected fields, and state resets. It stops short of saying what the call returns or whether the response echoes the resulting preference set, which is a minor gap given no output schema exists.

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 already 100%, so the baseline is 3; the description goes further by explaining patch semantics for omitted keys and the exact null-vs-true semantics for snoozeUntil, which the schema does not spell out. It does not walk through the other ten booleans, but those are self-describing in the schema.

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 name and title ('set_notification_prefs' / 设置我的通知偏好) plus the description's 'confirm then patch' framing and field-specific rules make clear this mutates the caller's own notification settings. It is distinguished from the read sibling by the closing '先get_notification_prefs', though the description itself never states the purpose in a single declarative sentence and relies on field names to convey scope.

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 routing: call get_notification_prefs first, leave SMS/email alone unless the user asks, and use different tools for per-group mute/leave and per-need pause, with set_my_match_preference named for avoid-person/direction settings. Both when-to-use and when-not-to-use are stated with named alternatives.

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