Agent Readiness · Checks for the person putting agents to work

Know your agents are ready before they act.

Putting agents to work brings a few practical questions: does each run leave records someone can follow from trigger to action, which money option fits a payment, do the rules cover an agent's first week, does what it can reach match what it may decide, does the evidence prove the change it reports, does a listing name its credentials before payment, can a retry pay twice, and does the kill switch stop everything. Agent Readiness answers them with eight tools: eight checks. Your agent sends the records, rules or design 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 readiness material and answers one question about it. The report names the record, rule or path behind every status, so your agent and you can see why it reads the way it does.

`wever_run-record-check`

Reads: One workflow run's records: the detection or trigger, the investigation notes, the approval and the actions taken, each with its id, links and time.
The signed report states: Whether the run is auditable end to end, tracing each action through its investigation and approval back to the trigger, with every missing or unlinked record named and every timestamp that is missing, unreadable or out of order.
Price: 0.10 USDC per call
Verdicts: unprovablegappedauditable

`wever_money-fit-check`

Reads: A task your agent must pay for, with its amount, counterparty and refund and liability needs, and the money options it holds: stablecoins, tokenized deposits, agent cards and others, with their stated properties.
The signed report states: The option that fits and why, each option ruled out by the property that rules it out, such as a final payment where the task needs a refund path, redemption risk or who carries loss, and the properties the options leave unstated.
Price: 0.10 USDC per call
Verdicts: no_option_fitsfit_unconfirmedfit_found

`wever_adoption-readiness-check`

Reads: The rules you plan to hand an agent: spending limits, approval gates, the accountable owner, recourse and refund handling, and logging, as fields or in your own words.
The signed report states: Each area as ready, thin or missing against the situations an agent meets in its first week, such as an unanswered approval or a refund request, with the smallest addition that covers each one.
Price: 0.10 USDC per call
Verdicts: not_readyready_with_gapsready

`wever_authority-scope-check`

Reads: What your agent can reach, system by system, and what it is authorized to decide: the actions it may take, their limits and approvers, and the data it may use.
The signed report states: Every action and data set the agent can reach outside its grant, including broad admin or write access, every granted action its access cannot perform, and the scoping change that brings each into line.
Price: 0.10 USDC per call
Verdicts: access_exceeds_authorityauthority_exceeds_accessaligned

`wever_change-proof-check`

Reads: A coding agent's account of what it changed and why, with the diff summary, the test results and the spec references it cites.
The signed report states: Whether the evidence proves the claim: each change with no diff or no passing test that covers it, each test cited for a change it does not exercise, each spec item left open and each diff entry no claim explains.
Price: 0.10 USDC per call
Verdicts: unprovenpartly_provenproven

`wever_credential-requirements-check`

Reads: One tool or service listing and its purchase flow, step by step, with the step where payment happens.
The signed report states: Whether your agent learns every API key, token, account, authorization, approval or plan before it pays, each one first disclosed at or after payment, and the place in the listing to state it.
Price: 0.10 USDC per call
Verdicts: disclosed_laterequirements_unnameddisclosed_before_paymentno_requirements_found

`wever_retry-safety-check`

Reads: A payment or refund flow's retry design, path by path: timeouts, retry triggers, idempotency keys, duplicate detection and how settlement is confirmed.
The signed report states: Every path where a retry can produce a second charge, refund or settlement, the paths an idempotency key, a status check or duplicate detection protects, and the smallest fix for each open path.
Price: 0.10 USDC per call
Verdicts: double_effect_possibleprotection_unverifiedretry_safe

`wever_kill-switch-check`

Reads: An agent system's kill-switch design: what fires it, what it stops, what happens to in-flight and queued work, who restarts it and how, with the system's parts and any drill results.
The signed report states: Whether the switch halts the system, with every part and every in-flight or queued action it leaves running, every restart path that resumes around it, and what the drills show.
Price: 0.10 USDC per call
Verdicts: label_not_controlhalt_unprovenhalts

How agents and people use it.

Your agent

Keeps the material each check reads: run records with their links and times, the money options it pays with, the rules it works under, its access list and authority grant, the evidence behind each change, the purchase flows it follows, its retry design and its kill-switch design.
Calls the check that answers the question on https://weverlabs.com/mcp at the moment it matters: when a run closes, before it pays, before it goes live, when its access changes, before it reports a change done, before it buys a tool, before a payment path ships and after a switch drill.
Acts on the verdict and keeps the signed report with the work, so the next step starts from a record.

You

See every run traced from trigger to action, with each gap named.
Know which payment paths are protected against a second charge or refund, and the fix for the rest.
Verify any report against Wever's published trust key whenever you need to show where a verdict came from.

A worked example

The refund retries on timeout. Its key changes on every attempt.

The request

{
  "name": "wever_retry-safety-check",
  "arguments": {
    "paths": [
      {
        "name": "Refund after support approval",
        "operation": "refund",
        "timeout": "10s",
        "upstream_completes_within": "30s",
        "retry_on": [
          "timeout"
        ],
        "max_retries": 3,
        "idempotency_key": {
          "present": true,
          "scope": "per_attempt"
        }
      },
      {
        "name": "Card charge at checkout",
        "operation": "charge",
        "timeout": "15s",
        "retry_on": [
          "timeout",
          "server_error"
        ],
        "max_retries": 2,
        "idempotency_key": {
          "present": true,
          "scope": "per_operation",
          "source": "order id"
        }
      }
    ]
  }
}

What the report returns

Verdict: double_effect_possible.

Refund after support approval	A retry after timeout can produce a second refund. The idempotency key is created per attempt, so each retry looks like a new request.
Card charge at checkout	A retry cannot produce a second charge: idempotency key holds across every retry.
Fix, Refund after support approval	Create the idempotency key once per refund, before the first attempt, and reuse it on every retry.
Fix, Refund after support approval	Before retrying a timed-out refund, look it up by reference and retry only when no refund exists.

1 path where a retry can produce a second charge, refund or settlement (Refund after support approval); 1 protected.

The refund path gives up after 10 seconds while the provider can take 30, so a retry starts while the first refund is still completing. Its idempotency key is new on every attempt, so the provider reads each retry as a fresh refund. The checkout charge reuses one key per order, so its retries settle once. One key per refund, and a status lookup before each retry, closes the refund path.

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 Readiness contract lists the schemas, limits, verdicts and signing details. For live order records, Job Records; for action gates and approval replays, Agent Oversight; for approver bands and spending policy controls, Agent preflight; for payment rails, Agent Treasury.

Browse all tools