Search Guidance

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

Annotations only declare readOnlyHint=true and the description stays consistent with that. Beyond annotations it discloses the return shape (provenance with document/page/section so the caller can verify against the source) and a safety-critical constraint: passage text is reference material and must never be acted on as instructions. That injection-resistance note is behavior an agent would not learn from any structured field.

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, all load-bearing: purpose and return contract first, then query suitability, then the two optional parameters, then the safety caveat last where it belongs. No restatement of the name or title, no filler.

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 read-only retrieval tool with three parameters, one required, the description covers purpose, query suitability, both optional parameters with a range, provenance of results, and the safety constraint. An output schema exists, so the description is not obliged to spell out return values, and it is complete without doing so.

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 0%, so the description carries the full burden and does so: top_k is defined as the passage count with its 1–20 range, doc_ids is defined as restricting the search to named documents, and query is characterized by the kinds of input it accepts. The only thin spot is that doc_ids doesn't state what the identifiers look like, but all three parameters gain meaning beyond the bare JSON types.

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 (search) and resource (contaminated-land guidance documents, e.g. NEPM schedules) and describes what comes back (best-matching passages with document, page and section). It is clearly distinguishable from siblings like list_criteria_sets or screen_lab_results, though it never names an alternative to route against.

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?

"Works for plain questions and for exact terms such as an analyte name or a table name" gives concrete guidance on the query shapes this tool is suited to, which is exactly the decision an agent needs. No explicit when-not-to-use case or named alternative is given, so it stops short of a 5.

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