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

Annotations cover the safety profile, yet the description still adds substantial behaviour the agent cannot infer: processing locality (AIProcurement.club), no retention, no Gateway memory, no external inference. It also discloses OCR behaviour ('auto handles empty-text pages, always handles mixed image/text') and an untrusted-data/injection warning, which is genuinely useful for a document-ingesting tool.

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?

Purpose is front-loaded in the first sentence and the constraint list follows in a compact second block with no filler. The middle section degrades into dense telegraphic fragments ('no Gateway memory, external inference, invoice or idempotency key'), which are efficient but harder to parse than a short structured list.

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?

Coverage of inputs, limits, auth and data handling is unusually thorough for a 3-parameter tool. With no output schema, however, the description never says what comes back — matched snippets, page references, per-file results, or the no-match case — which is the main thing an agent still cannot infer.

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 load, and it mostly does: it constrains orderId to an exact case-insensitive ID, discloses capacity ceilings beyond the schema (512 KiB combined, 12 pages/PDF, 20 pages/pack, 32k chars/document), and explains what 'auto' and 'always' do for the ocr enum. It leaves 'never' undefined and says nothing about the file.id/name/contentBase64 fields, so it is strong but incomplete.

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?

Names a specific verb and resource ('Find explicit labelled PO/Order references in supplied files') and pins the matching rule to exact ID, case-insensitive. It also rules out adjacent behaviours ('No fuzzy ID or filename joins'), which implicitly separates it from the search/extract siblings, though it never names a sibling tool outright.

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?

States the operating context clearly (agent-supplied files plus an order ID, exact match only) and adds a hard prerequisite, 'Requires agent:read'. It stops short of routing guidance — it does not say to use procurement_document__search when the ID is unknown or procurement_purchase_order__evidence when broader evidence is needed.

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