research

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

With no annotations provided, the description fully carries behavioral disclosure. It reveals the stealth browser mechanism, provenance receipts, stripped injection payloads, partial failure handling (source failures reported in `sources` rather than failing the whole call), per-domain rate limiting, and that it is the slowest sibling. It also mentions that failures are non-fatal and that sources are clamped. This is rich, honest disclosure beyond what structured fields could convey.

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?

The description is three short paragraphs, each deliberately front-loaded. The first sentence states the core purpose, the second explains the combined workflow and output, and the third gives usage guidance. Every sentence adds distinct information with no fluff or repetition. It is concise while remaining comprehensive.

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?

The tool is complex (combines search and page reading), has an output schema, and has no annotations. The description covers the key behavioral aspects: what gets returned (ranked passages with provenance and injection payloads), error handling (failures reported in `sources`), performance characteristics (rate limiting, slowest), and parameter effects (clamping, token budget). This is sufficient for an agent to correctly invoke and interpret the tool.

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 100%, but the description adds meaningful behavior for `max_sources`: values outside 1-10 are clamped, each source is a full page fetch, and it is the main cost/latency control. It also ties `query` to ranking and `max_sources` to the number of pages read. The description does not restate the schema, but adds the crucial clamping and cost semantics, going beyond the baseline.

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 and resource: 'Search the web and return ranked passages drawn from several sources.' It distinguishes itself from siblings by positioning as an alternative to `search` + repeated `read_url`, and contrasts with `search` (returns links only) and `read_url` (for already-known URLs). This clearly defines the tool's unique purpose.

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?

The final sentence gives explicit guidance: 'Prefer `read_url` when you already have the URL, and `search` when you only want links.' It also states when to avoid this tool ('the one to avoid for a single known page') and explains that it reads `max_sources` pages with rate limiting, making it the slowest of the three. This provides direct when-to-use/when-not-to-use direction.

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