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

Annotations cover the safety profile (readOnlyHint=false, openWorldHint=true, destructiveHint=false), and the description adds genuinely useful behavior: the job-ID return value, the required X-Api-Key header, BFS traversal order, and the non-obvious rule that a non-null wait_config auto-enables JS rendering while render_js=false plus wait_config is rejected. Some of that (auto-enable, rejection) restates schema prose, but the auth and async-job traits are description-only.

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 capability summary, then the wait_config nuance, then the return/auth facts. The wait_config sentences are dense and slightly over-explained relative to everything else, but every sentence carries information.

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 long-running, authenticated, open-world crawl job with no output schema, the description supplies the missing pieces an agent needs: what is returned (a job ID), how to retrieve results, and how credentials are passed. Deeper result-shape details are correctly deferred to crawl_status.

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?

Although the reported coverage metric is 0%, the nested schema is densely documented; the description still adds value by explaining depth as BFS levels and limit as a page cap in plain terms, and by spelling out the tri-state wait_config (omitted/null = adaptive settle, {} or concrete = explicit mode). It does not cover proxy geo-targeting defaults or formats in the same detail, so it is strong but not exhaustive.

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?

Opens with a specific verb+resource ('Start a website crawl job from a seed URL') and immediately distinguishes itself from the sibling 'crawl_status' by positioning this as the job-creating call and that as the polling call. No ambiguity about what the tool produces.

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?

Explicitly tells the agent to poll results with 'crawl_status' and states the authentication prerequisite. It does not contrast against nearby alternatives like 'batch', 'map', or 'extract', so the agent must infer when a single-site crawl beats a batch or map job.

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