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

Annotations (destructiveHint=true, openWorldHint=true, readOnlyHint=false) set the safety profile, and the description adds real behavioral context beyond them: first start installs Chromium and takes about a minute, stop ends the browser, get issues a fresh cdpUrl, and the in-sandbox endpoint is http://127.0.0.1:9222. 'stop ends it' also concretizes what the destructive hint means in practice.

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 purpose, then action semantics, then the install-latency caveat and the internal address. It is dense and telegraphic, with essentially no filler, though the run-on grouping of actions makes it slightly harder to scan than a bulleted breakdown would be.

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 carries return values (start returns cdpUrl, get returns state plus fresh cdpUrl) and the first-call install delay, which is exactly what an agent needs before invoking. The unqualified 'id' parameter is the remaining incompleteness.

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 60%; the description compensates by explaining the action enum values and the headless flag's real-world effect (visible window on the sandbox desktop with live view), while width/height retain their schema defaults. The main gap is that 'id' is never qualified as the sandbox identifier, leaving one required parameter unanchored.

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 concrete verb+resource ('Chromium in a sandbox for browsing agents') and then enumerates the three actions (start/get/stop) with what each does, so an agent can tell this apart from siblings like runtime_sandboxes_exec or runtime_sandboxes_desktop_screenshot 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?

Clearly signals the use case: obtaining a cdpUrl to hand to Playwright connectOverCDP, Puppeteer, browser-use or Stagehand, plus when to use headless=false (live desktop view). It lacks any explicit when-not or comparison against the desktop_* siblings, so it's context-rich but not fully routing.

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