Evaluation drift check
wever_evaluation-drift-checkAgent Evaluation: submit two evaluation runs of the same agent at different times, each with its id, date, model and model version, prompt version, dataset and dataset version, judge and judge version, rubric version, temperature and scores (each with its metric, value, sample size and spread), and optionally the tolerance a score may move by. The signed report compares every condition between the runs, the same, changed or not stated, and every score with how far it moved against the tolerance or, with none, the two runs’ combined noise, and names every changed model, prompt, dataset, judge, rubric or setting that makes the runs measure different things, then every score that drifted, every metric scored in one run only and every score on less than half the earlier sample, with the fix for each. This compares two runs of one agent: whether an implementation’s behavior matches its spec is wever_spec-drift-check, a live endpoint’s reachability, latency and consistency is wever_endpoint-reliability-check and what one run needs to be rerun is wever_evaluation-reproducibility-check. Verdict runs_not_comparable, scores_drifted, drift_unverified or scores_stable. POST /api/evaluation-drift-check. Reads only the material in the call: no network request, fetch, money movement, message, sharing, configuration change or external write. Paid URL: 0.10 USDC per call on Base via x402 exact, with the shared 10 signed-wallet free calls per rolling 30 days. Bind X-Wever-Wallet-Proof to the exact JSON arguments and the tool route. Returns a report with an Ed25519 signed receipt. Contract: /api/agent-evaluation-contract. Human page: https://weverlabs.com/agent-evaluation/.
Tool call template
Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.
{
"id": 1,
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"arguments": {
"earlier": "<earlier>",
"later": "<later>"
},
"name": "wever_evaluation-drift-check"
}
}Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| agent | No | The agent both runs evaluate. | |
| label | No | Optional name recorded in the report. | |
| later | Yes | The later run. | |
| earlier | Yes | The earlier run. | |
| tolerance | No | How far a score may move before it counts as drift, such as 2. |