Run an MCP tool call in the VM

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

Beyond the annotations (readOnlyHint=false, openWorldHint=true, idempotentHint=false), the description discloses a concrete behavioral trait: 'Evidence is automatic: snapshots before and after, and a step record,' and that reason/expected/afterIntervalMs only override that evidence. The caution 'Never replay an uncertain call' aligns with the non-idempotent, open-world hints. It stops short of a 5 because it does not address failure/error propagation or side-effect expectations.

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?

Four sentences, roughly 70 words, each earning its place: purpose, target enumeration, discovery routing, evidence model, and a safety caution. The core verb-resource pair is front-loaded in the first sentence, and there is no fluff or redundant restatement of schema descriptions.

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 complex 7-parameter open-world tool with no output schema, the description covers the essential operational knowledge: target choices, forwarding semantics, the discovery prerequisite, automatic evidence, and optional overrides. Minor gaps remain — the exact return/step-record shape and the lease prerequisite implied by siblings relay_acquire/relay_release — but these do not prevent a correct invocation.

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 100%, so the baseline is 3. The description adds meaning by grouping parameters into two semantic classes — tool/args are 'forwarded unchanged' (the actual remote call) versus reason/expected/afterIntervalMs as 'optional overrides' of the evidence record — a relationship the per-property schema entries only imply. This framing tells the agent which parameters affect the remote server and which only affect the step record.

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 opens with a specific verb-resource pair, 'Send one tool call to an MCP server inside the VM,' and enumerates the three valid targets with parenthetical mappings (cua, playwright, chrome-devtools). It also clarifies the forwarding semantics for tool and args, and the relay_tools reference helps distinguish this tool from its discovery sibling without opening the schema.

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?

It explicitly routes the discovery step to the relevant sibling — 'Use relay_tools to see a target's exact tools' — which is the natural confusion point. It also states a when-not condition with 'Never replay an uncertain call' and clarifies which parameters are optional evidence overrides, so an agent knows what to supply and what to leave out.

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