run_command

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

Annotations only declare non-readOnly/destructive/openWorld; the description goes well beyond by enumerating the outcome taxonomy (normal | exception | silence | timeout | checksum_error | malformed), explaining `expect` auditing and that unmet expectations return an error with full result JSON, and disclosing the risk gate (acknowledge_risk=true for risk=danger). That is substantive behavioral context an agent cannot get from the annotations.

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?

Front-loaded with the action, and every clause carries information about outcomes, session model, or risk gating. The middle sentence packs session_id, port, selector omission, and replay syntax into one long clause chain, which is dense though not wasteful.

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 destructive, open-world tool with 7 params, a nested object, no output schema, and low schema coverage, the description supplies the outcome model, audit/error semantics, session-vs-ephemeral choice, and the risk acknowledgement gate. It stops short of describing how per-command `params` are shaped or how results are summarized by max_inline_bytes beyond the schema's own note.

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 29%, so the description has to compensate, and it largely does: session_id vs port semantics, the selector-based port omission rule, the replay:<capture path> syntax, and the acknowledge_risk requirement are all description-only. The nested `params` object and `device`/`command` remain only lightly characterized ('typed params'), leaving one meaningful gap.

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 ('Execute a named command from devices/<device>/commands/') and immediately explains the mechanism (builds the frame, awaits response, classifies outcome). This is clearly distinct from lower-level siblings like send/request or the orchestration tool run_workflow, even though no sibling is named.

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?

Gives real usage context: use session_id for an existing session, or port for an ephemeral transport, with port optionally omitted when the profile declares a selector, and replay:<capture path> accepted. It does not, however, tell the agent when to prefer this over alternatives such as run_workflow or send, so there is no explicit when-not/alternative guidance.

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