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

Annotations already declare readOnly=false, idempotent=true, destructive=false, and the description adds substantially: one export debits billing.export.asset_period, failed generation reverses the debit, refusal refunds and carries no rows, unpaid calls return a structured payment_required challenge with no fabricated data, and other payers' private subjects are redacted to null. This is rich behavioral context beyond the annotations.

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?

It leads with payment mechanics rather than the core purpose, and repeats 'After paying, poll /v1/resolutions/:id until horizon_end' and the PAID/USDC framing twice. Dense and information-rich, but several clauses could be cut without loss, so it is not well front-loaded.

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 paid, billing-affecting export tool an output schema already exists, so return values need not be described; the description instead covers pricing, payment rails, idempotency/replay, refunds, redaction, and challenge-on-unpaid behavior. An agent has everything needed to invoke it correctly.

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?

With 0% schema description coverage the description must carry the load, and it explains payer (tool arg, x-tollbridge-payer header, or x402 PAYMENT-SIGNATURE), scope (asset symbol vs question id), subject, and format (CSV/JSONL). It only indirectly implies the from/to window via 'one UTC calendar month' and 'oversized window', leaving the two date-time parameters unlabeled.

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+resource: export the decision history for one subject over one UTC calendar month as CSV or JSONL, with an explicit field list (claim, distribution/probability, confidence, hashes, decided_at, horizon, outcome). An agent can distinguish it from the sibling read tools (tollbridge_decision, tollbridge_get_decision) without opening the schema.

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 operational context: it is a paid add-on, the same subject+month is a typed replay that is never re-charged, an oversized window is refused and should be narrowed, and exact price comes from tollbridge_pricing. It does not, however, route the agent to or away from specific sibling tools for the same data, so it stops short of explicit alternatives/exclusions.

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