adl_gesture

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

With no annotations, the description carries the disclosure burden and does so well: frame freshness bound (max 5s), optimistic-concurrency revision, point-count and timing constraints, pre/mid-motion safety checks, and the return contract (released-input acknowledgement plus fresh PNG, with application success explicitly not inferred). It stops short of permissions/auth and failure/error semantics, which keeps it below a 5.

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?

Purpose is correctly front-loaded in the first clause and no sentence is purely filler. But the text is telegraphic and compressed to the point of ambiguity — 'age5s', 'and2–64', 'first0' — which costs readability on constraints an agent must parse exactly.

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?

This is a high-complexity, 7-parameter, no-annotation, no-output-schema tool, and the description still covers preconditions, timing/point constraints, staleness rules, and the shape of the response. The remaining gaps are session/surface identity scope and error behavior, which are meaningful but secondary.

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 0%, so the description must do the work. It adds real meaning for frame_id (latest, max age 5s), expected_revision, and the points array (2–64 CSS/window-pixel samples, t_ms starting at 0 and increasing with total <=5000). But session_id, surface_id, and swipe_backend are left entirely to the schema, leaving roughly half the parameters unexplained.

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?

States a specific action pair (drag/swipe) against a defined resource (a trusted surface), and the kind enum is echoed in prose so the agent knows the two modes. It differentiates context (browser vs interactive_desktop, scroll_swipe vs touch_swipe) but never names a sibling tool, so sibling disambiguation is only partial.

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 description explains backend choice (isolated mouse/genuine touch vs physical OS input, scroll vs touch swipe) and lists preconditions (target/focus/human-input checks), which implies when the call is valid. However it never says when to prefer this over adl_act, adl_execute or adl_prepare/commit_input, so the agent gets no explicit sibling routing.

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