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

Annotations only declare readOnly/openWorld/idempotent/destructive hints; the description adds substantially more: the transport (POST /api/duplicate-effect-check), an explicit no-side-effect guarantee (no network request, fetch, money movement, message, sharing, configuration change or external write), the pricing model (0.10 USDC on Base via x402 exact, 10 free signed-wallet calls per rolling 30 days), the required X-Wever-Wallet-Proof bound to the exact JSON arguments and route, and the Ed25519-signed receipt. That is exactly the auth/pricing/side-effect context annotations cannot carry.

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 is front-loaded with purpose and does carry a usable structure (purpose, scope and siblings, verdicts, endpoint, safety, price, contract link). But it is a run-on of very long sentences and repeats the effect taxonomy ('a message, an order, a record, a booking or something else' and again 'messages sent twice, orders placed twice, records created twice, bookings doubled'), plus the guard list appears twice; trimming those redundancies would cost nothing.

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?

There is no output schema, and the description compensates by describing the report contents (effect-by-effect verdict, unguarded effects, per-attempt idempotency keys, effects seen happening more often than expected, repeats only cleaned up afterward, the fix for each) and the verdict vocabulary. Pricing, auth binding, and the contract/human-page links round out what an agent needs before calling.

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 100%, so the schema already documents flow, runs, label and every effects sub-field including the guard and key_per enums. The description restates those semantics (guard variants, what one key covers, checked-after-recovery, receiver-drops-repeats) rather than adding syntax, defaults, or edge-case rules beyond the schema — baseline 3 applies.

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: it checks, effect by effect, whether duplicated effects are caught after a resume/retry, and emits a signed report with a named verdict (effects_can_repeat, duplicate_gaps, duplicate_unverified, duplicates_checked). It also explicitly carves out the payment case and names the sibling that owns it, so an agent can tell it apart from wever_double-charge-check, wever_duplicate-order-check and wever_retry-safety-check without opening a 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 explicit routing rules: duplicated payments go to wever_double-charge-check, one timed-out checkout against merchant order records to wever_duplicate-order-check, and retry safety of a payment/refund flow to wever_retry-safety-check. The exclusion ('every effect a task can duplicate other than payment') and the positive case (effects a resumed or retried task can repeat) are both stated.

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