adb_shell

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

No annotations are provided, so the description carries the full burden of behavioral disclosure. It fails to mention that this tool can execute arbitrary, potentially destructive commands (rm, reboot, etc.) with no confirmation, or that some devices may require root. The 'use only with devices you own' line is an ethics note, not a behavioral transparency statement about side effects or permissions.

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 description is a single sentence plus a one-line ethical caveat. It's efficiently brief with no filler words. The ethical warning is a legitimate addition worth including, and the entire thing is properly front-loaded.

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?

Given zero annotations and no output schema, this is a serious gap. This is a high-risk arbitrary-execution tool — arguably the highest-risk operation in the adb family — yet it lacks any behavioral detail about output capture, exit codes, error handling, or the fact that commands run with shell (often root) privileges. Sibling tools like adb_getprop, adb_settings, adb_logcat already cover structured queries, and the description should note this tool is the general-purpose escape hatch.

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 only 50% — the 'command' parameter is entirely undocumented in the schema (no description field). The description says 'shell command' which clarifies command's nature somewhat, but adds no detail about what commands are supported, shell language (sh, bash?), or escaping requirements. The serial parameter is well documented in the schema already.

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 states it runs a shell command on the device, which is a clear verb+resource pairing. However, it doesn't differentiate from siblings like adb_getprop, adb_settings, or adb_logcat, which are more specialized forms of reading device state. The broad nature of 'shell command' is clear but the description doesn't scope what kinds of commands are appropriate.

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 only usage guidance is 'Use only with devices that you own' — an ownership/ethics constraint, not operational guidance. There's no when-to-use versus alternatives (e.g., when to use adb_getprop vs adb_shell for property queries), no mention that many sibling tools cover common shell operations, and no safety warnings about destructive commands.

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