previewapk_act

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

With no annotations, the description carries the full burden and does so: it discloses ordering, the 20-action cap, non-atomic execution ('stops at the first failure — earlier actions stay done'), and the response shape ('per-action results with timing and error envelopes'). The non-atomicity and no-blind-replay rules are exactly the behavioral traits an agent needs before invoking a mutating batch tool.

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?

The size is justified by the action vocabulary that would otherwise be undocumented, and the batch constraint plus failure semantics are front-loaded. Structure is weak, though — a single long run-on sentence of comma-separated enumerations makes it harder to scan than a short bulleted action list would be.

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 20-item mutating batch tool with no annotations and no output schema, the description covers the essentials: allowed actions, ordering, failure behavior, response envelope at a summary level, and post-batch observation. Missing pieces are the `device` parameter and any note on whether a failed batch can be resumed or selectively re-run.

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% and `actions` is a bare array of untyped objects, so the description must compensate — and it largely does, spelling out the payload shape for tap (x/y fractions, text/id/desc with nth, or ref+snapshot_id), swipe, long_press, double_tap, type, key, open_app, close_app, open_url, notifications, wait, and wait_for. The gap: the `device` parameter is never mentioned, and observe is described only for 'ui'/'screenshot'/'both' while the enum also allows 'none'.

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?

Opens with a specific verb+resource+scope: 'Run a batch of up to 20 actions in order on the device,' then enumerates every supported action type with its payload shape. The batch framing implicitly but clearly distinguishes it from the singular siblings (previewapk_tap, previewapk_swipe, previewapk_type_text, previewapk_press_key), so an agent can pick it without reading schemas.

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?

'NEVER replay a batch whose outcome you don't know; observe first' is an explicit when/when-not rule, and the observe parameter guidance tells the agent to attach the screen after the batch. It stops short of stating when to prefer this tool over the one-action siblings (e.g., multi-step sequences vs. a single tap with immediate verification).

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