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

Goes well past the readOnly/openWorld/idempotent annotations: discloses the billing model (unverifiable rows free, fills billed 1.0), that confidence is a routing signal not an accuracy claim, that unmapped labels are never best-effort matched, and that disagreements produce a dispute record with a review SLA. These are exactly the operational traits an agent needs before invoking a metered verification 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?

Front-loaded with the verb and the core mechanism, but the middle becomes a long run-on enumeration of every return field (verdict, cte_category, cte_treatment, cte_taxable, confidence, evidence, suggested_fill, dispute_id, failed check, review_sla_hours). Since an output schema exists, that inventory is largely redundant and crowds out the higher-value routing and billing statements.

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 batched, metered verification tool the description covers mechanism, vocabulary handling, billing consequences, failure modes, and hard limits, and the return shape is backed by an output schema. Nothing an agent needs in order to decide whether and how to call it is missing.

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%, so the baseline is 3, but the description adds real meaning beyond the schema: the top-level fill flag is a default for rows that don't set their own, vocabulary resolution order (Standard first, collisions reported), and the label_unmapped outcome for unrecognized labels. Only the row-level fields like wallet perspective and chain auto-detection are left entirely to the schema.

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 and resource: verify a batch of caller-assigned labels against a deterministic classification basis (rules, verified cache, answer guards) with no model call. The 'no model call' phrase and the fill=true escape hatch implicitly separate it from classify_batch, so an agent can tell which tool classifies versus which adjudicates a caller's label.

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?

Gives clear routing rules: a row without deterministic basis is unverifiable and free, and passing fill=true runs the classification engine and bills that row as a fill. It also states scope limits ('does not decide tax treatment beyond the published Standard and does not file'). It never names classify_batch or classify_transaction as the alternative for pure classification, so the when-not guidance is implied rather than explicit.

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