act

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

With no annotations, the description does real work: it discloses the return shape ('effect', a plain sentence describing navigation/dialog/scroll/no-change, plus snapshot_id) and the observe toggle. It stops short of warning about irreversible side effects (clear wipes text, taps may trigger destructive flows) or timeout behavior.

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?

Front-loaded with the combined act+observe premise, then targeting, then return value, then the observe escape hatch — a logical order with little waste. The opening 'One human-shaped verb' is slightly rhetorical, and the dense parenthetical list of selector fields is a bit compressed.

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?

Covers the essentials for a high-traffic interaction tool: what it does, how to target, what comes back, and the no-observe fast path. Gaps remain for the scroll/key-specific params (direction, duration_ms) and the wait_timeout that governs how long it blocks.

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 description coverage is 0% across 9 params, so the description must compensate. It well documents `action`, the selector predicates (role, label_contains, semantic_key, text, resource_id) and `observe`, but leaves `value`, `direction`, `duration_ms`, and `wait_timeout` entirely 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 concrete verb and effect: 'perform an interaction AND report what changed, in a single round trip,' then enumerates the action space (tap/long_click/set_text/.../key). It implicitly separates itself from the pure-inspect siblings (ui_snapshot, get_ui_tree) by saying no follow-up perceive call is needed.

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?

Gives a clear condition for the alternate mode ('Set observe=false for tight loops where you will observe later yourself') and explains the two targeting routes (semantic selector vs x/y). It does not explicitly say when to prefer sibling tools like tap_element, input_text, or element_action, which leaves some routing ambiguity.

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