Preview what NVDA would say

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

Beyond the readOnly/idempotent/non-destructive annotations, it discloses that the scan does not move focus or press keys, that it flags unlabeled or silent controls, and — most usefully — that returned controls get ids consumable by nvda_focus_object, nvda_activate_object, and nvda_get_dev_info. That cross-tool handoff is behavioral context annotations cannot convey.

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 tight paragraphs: the core action and its non-effects come first, the mode behavior and sibling routing follow. Every clause carries information — the parenthetical tool lists are the only density cost, and they are load-bearing for chaining.

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 an output schema present, annotations covering the safety profile, and 100% schema coverage on all four parameters, the description only needs to supply routing and chaining context — which it does fully. An agent could call this correctly on the first attempt.

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% with per-parameter descriptions, so the baseline is 3; the description adds genuine meaning by explaining that 'window' includes controls Tab cannot reach and that the scan happens 'at once,' which the schema does not state. It does not add format or interplay detail beyond that.

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 gives a specific verb and resource ('Say what NVDA would announce when each control gets focus') plus two concrete side-behaviors (no focus movement, flag unlabeled/silent controls). It also explicitly contrasts itself with nvda_audit_focus_order, so an agent can separate it from siblings without reading either schema.

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?

It names the alternative tool explicitly and the condition that selects it: 'to test what happens when keys are pressed, use nvda_audit_focus_order.' It also scopes the 'window' mode (whole-window scan including controls Tab cannot reach), leaving no ambiguity about when each mode applies.

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