shell_run

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

Annotations already declare destructiveHint=true, readOnlyHint=false, and openWorldHint=true, so the safety profile is covered. The description adds a process-level behavioral detail beyond annotations: invoking this tool triggers the same risk gate as host_run, and it openly labels the tool as an arbitrary-command escape hatch. It stops short of explaining what the risk gate entails, but the added context is meaningful.

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 front-loaded with the core purpose, followed by risk context, alternative guidance, then args/returns. The Args block partly restates schema content, but the host-field warning earns its place; overall the structure is logical with no filler.

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 high-risk arbitrary-command tool, this covers purpose, the risk-gate behavior, alternative routing, and the host-field trap. The output schema documents return formats and annotations cover the safety profile, so nothing critical is missing. A short note on what the risk gate actually does would push this to fully complete.

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?

The schema already documents every parameter (command, sudo, timeout, cwd, response_format, host) with descriptions, so the baseline is met. The description adds the single most critical semantic warning: leave 'host' empty or the command will run on a remote machine — a trap an agent would not infer from the schema alone. Minor blemish: it names the type 'HostRunInput' while the schema defines 'RunInput', but the listed fields align.

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 opening line states a specific verb, resource, and scope: run a one-time shell command on the LOCAL machine where the MCP server runs. This immediately distinguishes it from host_run (remote) and from the specialized system_* and apt_* siblings, and the 'escape hatch' framing reinforces its role as a general-purpose fallback.

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 description gives explicit routing guidance: prefer the specialized tools (system_*, apt_*, etc.) and use this only as an escape hatch for custom commands. It also warns that the same risk gate applies as for host_run and that filling the 'host' field redirects execution to a remote machine, clarifying when this tool is versus is not appropriate.

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