Agent Purchases · Checks for the buying side of paid agent calls

Know what your agent bought.

An agent that pays for a call needs the price before it pays, a result that matches the order, one charge per call and a way back when a call fails. Agent Purchases checks each of those with eight tools: eight checks. They cover whether a provider is ready for agent buyers, what a rehearsal proves before money moves, whether the charge matches the quote, whether the result is what was ordered, whether a call was charged twice, whether a refund route actually works, whether the buyer identity on a receipt is properly delegated, and whether the buyer's own records can show a failure at all. Your agent sends the facts, quote, result, call record, route, receipt or ledger in the call, and each tool returns a signed report with every status, match and verdict decided from it.

See the tools Read the contract

Tool access: 0.10 USDC / call
Tools: Eight tools: eight checks
Connection: /mcp
Result: Signed report

The eight tools.

Each tool reads one kind of purchase material and answers one question about it. The report names the fact, dimension, field, term, settlement, step or record behind every status, so your agent and you can see why it reads the way it does.

`wever_provider-readiness-check`

Reads: A provider's public purchase facts as an agent buyer meets them: what it discloses and when, how a call is priced and quoted, the rehearsal it offers and how money moves back.
The signed report states: Each readiness piece (disclosure before payment, price and quote, rehearsal, money back) as ready, implied or missing, with every fact shown only after payment, every quote that does not bind and every missing or contact-only refund route.
Price: 0.10 USDC per call
Verdicts: not_readyimplied_readinessreadiness_unverifiedready_for_agent_buyers

`wever_rehearsal-check`

Reads: One paid call and every dry run, sandbox, simulated response or test call the provider offers for it, with the price each shows, the payment step it runs, what comes back, its speed, its failures and its network.
The signed report states: Which rehearsal proves the price, the payment handshake, the result shape, timing, failure paths and the network, why each other falls short, and every dimension the first paid call would be the first to test.
Price: 0.10 USDC per call
Verdicts: no_rehearsalrehearsal_gapsrehearsal_unverifiedrehearsal_covers_call

`wever_price-truth-check`

Reads: The quote an agent received for one paid call, with its amount, asset, network, quantity and expiry, beside the charge settled for it.
The signed report states: Every mismatch in amount, asset, network or quantity, with an amount that moved between quote and settlement given with its difference and share of the quote, units charged but not delivered, and a charge settled after the quote lapsed.
Price: 0.10 USDC per call
Verdicts: price_mismatchquote_gapsprice_unverifiedprice_true

`wever_result-validation-check`

Reads: The order's own terms for one paid call (the values it asked for, the fields it requires, how fresh the data must be, the fewest items that count) beside the result that came back.
The signed report states: Each term met or not met, with every wrong value, missing field, error inside a success, empty result and stale result. A 200 carrying the wrong data is a silent failure.
Price: 0.10 USDC per call
Verdicts: silent_failureorder_not_metvalidation_unverifiedorder_met

`wever_double-charge-check`

Reads: The calls a buyer ordered once and the call sequence as recorded: every 402 challenge, settlement, refund, delivery and failure, with times, amounts and settlement references.
The signed report states: Every call settled more than once with each reference and the extra amount standing, every 402 returned after payment asking to be paid again, every settlement for a call never ordered and every duplicate a refund answered.
Price: 0.10 USDC per call
Verdicts: double_chargedrepay_requestedcharges_unverifiedcharged_once

`wever_refund-path-check`

Reads: A provider's stated refund route for failed paid calls, each step with where it happens, what a test of it showed and what the buyer needs, plus the failed call and the proof the buyer holds.
The signed report states: Each step as reachable, unreachable, unpublished or untested, with every block on the way back (no route, contact-us alone, a closed window) and every gap (an account an agent may not hold, credit in place of money, proof the buyer lacks).
Price: 0.10 USDC per call
Verdicts: no_working_routeroute_gapsroute_unverifiedroute_reachable

`wever_delegated-identity-check`

Reads: One purchase receipt, with the buyer identity it names, the acting agent and the credential paid with, beside the delegation behind it: who delegated, to which agent, its scope, its expiry and how it is revoked.
The signed report states: Whether the receipt names the delegating business as the buyer and the delegated agent as the actor, inside the delegation's merchants, categories and amount, before expiry and revocable, or whether the agent is paying as itself.
Price: 0.10 USDC per call
Verdicts: pays_as_itselfnot_delegateddelegation_gapsidentity_unverifiedproperly_delegated

`wever_purchase-record-check`

Reads: The buyer's own record of paid calls: what was ordered, from whom, at what quote and amount, to which destination, with which reference, receipt and result.
The signed report states: What the ledger shows was paid, every missing amount, destination, result and receipt, and for each invisible failure (price drift, a double charge, a 200 with the wrong data, money owed back) how many records can show it.
Price: 0.10 USDC per call
Verdicts: failures_undetectablerecord_gapsreconstructable

Start with the two free tools.

On https://weverlabs.com/mcp-free, anonymous and with no wallet, two Agent Purchases tools read the same kind of material at summary depth. Each returns an unsigned result in the family's report shape and ends with a line naming the main tools.

`wever_provider-readiness-sample`

Reads: Up to ten purchase facts a provider shows a buying agent, each with when the agent meets it and whether it is stated or only implied.
Returns: For each fact, whether it is disclosed before payment. The provider readiness check gives the full audit.
Access: Free on /mcp-free

`wever_call-terms-snapshot`

Reads: A service's stated terms for one paid call: the amount and asset, the network, what counts as delivered and the window for money to move back.
Returns: Each of the four terms in plain words, or not stated, before your agent pays.
Access: Free on /mcp-free

How agents and people use it.

Your agent

Reads a service's terms first with wever_call-terms-snapshot, so it knows the amount, network, what counts as delivered and the window for money to move back before it pays.
Calls the check that answers the question on https://weverlabs.com/mcp at the moment it matters: before it first buys from a provider, before the first paid call, after each settlement, when a result comes back, when it reconciles its charges, when a call fails and money is owed back, when it pays under a business's identity and when it closes its books.
Acts on the verdict and keeps the signed report with the purchase, so the next step starts from a record.

You

See each gap named with the change that closes it.
Know which charges matched their quotes, which results matched their orders, which calls were charged twice and which refunds have a working route.
Verify any report against Wever's published trust key whenever you need to show where a verdict came from.

A worked example

A 200 response. The wrong data.

The request

{
  "name": "wever_result-validation-check",
  "arguments": {
    "order": {
      "call": "GET /v1/price",
      "requested": [
        {
          "field": "symbol",
          "value": "ETH"
        },
        {
          "field": "currency",
          "value": "USD"
        }
      ],
      "required_fields": [
        "price",
        "as_of"
      ],
      "max_age": "5 minutes"
    },
    "result": {
      "status": 200,
      "fields": [
        {
          "name": "symbol",
          "value": "ETH"
        },
        {
          "name": "currency",
          "value": "EUR"
        },
        {
          "name": "price",
          "value": "2874.10"
        },
        {
          "name": "as_of",
          "value": "2026-11-09T11:40:00Z"
        }
      ],
      "as_of": "2026-11-09T11:40:00Z",
      "received_at": "2026-11-09T12:00:00Z"
    }
  }
}

What the report returns

Verdict: silent_failure.

Failure, wrong value	The order asked for currency USD, and the result carries currency EUR.
Failure, stale result	The data was current at 2026-11-09T11:40:00.000Z and arrived at 2026-11-09T12:00:00.000Z, 20 minutes old, past the 5 minutes the order allows.
Fix, currency	Ask the provider for the result the order asked for: currency as requested.
Fix, stale result	Ask the provider for data within the freshness the order requires, or a refund.

The call returned 200, and the result fails the order: wrong value and stale result. A success status with the wrong result is a failure.

The agent ordered an ETH price in US dollars, no older than five minutes, and paid for it. The call returned 200 with a symbol, a price and a timestamp, so a status check would count it as delivered. The result carries the price in euros, and its data was twenty minutes old when it arrived. The report names both, so the agent asks for the result it paid for instead of acting on the wrong one.

Call it from your agent.

Connect to https://weverlabs.com/mcp and call any of the eight tools. Each call costs 0.10 USDC on Base. Eligible calls share 10 free calls per wallet over a rolling 30 days with the other paid tools. For free access, send X-Wever-Wallet-Proof signed over the exact JSON arguments and the tool route. The payment challenge supplies the x402 terms for a paid call.

Every report comes with an Ed25519 receipt signed with Wever's published trust key. The Agent Purchases contract lists the schemas, limits, verdicts and signing details. For retry designs and run records, Agent Readiness; for refund rules and identity portability, Agent Accountability; for terms matches and approval records, Agent Governance; for settlement proof and external listings, Agent Access.

Browse all tools