Set Device Power

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

The note that powering off a Cisco device discards its unsaved running-config is genuinely valuable behavioral context that the annotations do not convey. But it directly conflicts with destructiveHint=false, creating ambiguity about whether this call can lose data, and there is no statement about idempotent re-application (already-on/off), device-type scope, or persistence of state beyond the session.

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?

One sentence of purpose followed by a high-value caveat and two short examples – nothing extraneous, and the destructive side effect is front-loaded where an agent will read it. The examples are mildly redundant with the boolean parameter but still demonstrate the natural-language trigger phrasing.

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?

With a defined output schema, return values need no explanation, and the two required parameters are largely covered. What is missing is the safety framing (should you save first? does the warning apply only to Cisco devices?) and any statement of side effects on interfaces, which is the main residual risk for a state-changing tool.

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% (power_on has no description), and the description compensates by mapping the boolean to natural-language intent via the examples and the phrase "on or off". The name parameter's display-name semantics are already covered by the schema, so the added value sits almost entirely on power_on, which is the gap that mattered.

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?

"Power a device on or off" gives a specific verb (set power state) and a specific resource (a device), which is immediately actionable. It is implicitly distinguishable from the sibling set_interface_power, but the description never names that sibling or states the device-vs-interface scope explicitly, so the differentiation is left to inference.

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 two examples ("Turn off R1", "Power on all the switches") illustrate realistic invocation contexts, and the unsaved-config warning implies a prerequisite (save first). However, there are no explicit when-to-use/when-not rules and no routing to alternatives such as set_interface_power or save_device_configuration, leaving the agent to guess which sibling applies.

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