Fork sandbox

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

Goes well beyond the destructiveHint/readOnlyHint annotations: the source pauses momentarily but survives, each copy is separately billed and limit-counted, partial success is possible with one entry per copy ('a sandbox or an error'), and copies inherit the source lifetime unless lifetimeMs is set. That is exactly the mutation side-effect detail an agent needs.

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?

Five short sentences, all load-bearing, with the core action front-loaded and the safety gate (user approval) placed last as a final check. Dense but never padded; only slight cost is the packing of billing and limit behavior into one sentence.

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 no output schema, the description supplies the return shape (one entry per copy, sandbox or error), plus billing, lifetime, and limit semantics. Minor gaps remain – no cost guidance for large counts and no statement of what state a fork starts in – but an agent can invoke this correctly.

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 only 25%, so the description carries weight; it does explain lifetimeMs (inherits source lifetime unless given) and hints at count through 'one entry per copy' and 'only some may start'. However, count is never explicitly stated as the number of copies to create and confirm is only obliquely covered by 'Get the user's approval first', leaving half the parameters under-documented.

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 – 'Start live copies of a running sandbox: memory, processes and files' – which cleanly separates it from sandbox_snapshot (state capture) and sandbox_create (blank sandbox). The 'live copies ... memory, processes and files' phrasing tells an agent exactly what is duplicated.

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 clear preconditions: source must be running, the user must approve first, and copies count against running-sandbox limits so bulk forks may partially fail. It does not name the alternative tools (snapshot vs. create) an agent might pick instead, but the operational context is unambiguous.

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