adb_shell

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

With no annotations, the description carries the full burden and only discloses that output is returned. It says nothing about the fact that arbitrary shell commands can mutate device state, whether the device must be connected/unlocked, whether there is a timeout, or how failures are surfaced — all critical for a raw escape-hatch 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?

A single front-loaded sentence with no filler, which is appropriate for a one-parameter tool. The parenthetical '(advanced)' is the only soft spot — it's a vague qualifier that would be better spent on a concrete usage condition.

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 raw, unannotated, potentially destructive shell tool with an undocumented parameter and no output schema, the definition is far too thin. An agent can call it, but cannot predict safety, device preconditions, output format, or failure behavior — all of which this tool needs since nothing structured fills the gap.

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% for the single required 'command' parameter, and the description adds no meaning: it doesn't clarify whether the string is the full shell line including arguments, how quoting/escaping works, or whether multi-command sequences are allowed. The one sentence 'Run a raw adb shell command' essentially restates the parameter name.

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 verb and resource ('Run a raw `adb shell` command on the phone') and signals the return ('return its output'), which clearly separates it from the high-level primitive siblings like tap, swipe, and type_text. It stops short of 5 because 'raw' and '(advanced)' are not elaborated on, so the actual scope of what can be run stays implicit.

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 parenthetical '(advanced)' is the only hint that this is a fallback, and the description never says when to prefer adb_shell over the dedicated tap/type_text/open_app siblings or when to avoid it. There are no prerequisites, no device-state requirements, and no exclusions, leaving the agent to infer routing entirely.

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