Evaluation scoring check
wever_evaluation-scoring-checkAgent Evaluation: submit one scoring rubric and its judge setup: the rubric version, what it scores, the scale’s lowest and highest scores and step, and each scale point with its anchor and whether a worked example is attached; who scores (a model, people, an automatic metric or a mix) with the judge model, whether its version is pinned and its prompt fixed, its temperature, whether it is the model being scored, and its calibration (against human labels, known answers or nothing, on how many items and at what agreement); and the spot check, whether people check a share of the scores, what share, who checks and how often they disagree. The signed report says point by point whether the scale is anchored, whether a model judge is calibrated and pinned, whether a person spot-checks judged scores and whether scores compare across runs, and names an unanchored scale, an uncalibrated model judge, judged scores no person checks and an unpinned judge, then every unanchored point, points off the scale, no worked examples, an unversioned rubric, calibration agreement under 0.7 or on fewer than 50 items, an unfixed judge prompt, a sampling judge, a judge scoring itself, a spot check under one in twenty or by the builders and spot-check disagreement over 0.2, with the fix for each. This reads the rubric and judge behind a score: a trust score computed from verified completed work is wever_agent-trust-scorecard and whether a job’s delivered work meets its acceptance criteria is wever_deliverable-check. A worked rubric is wever_evaluation-rubric-sample on https://weverlabs.com/mcp-free. Verdict scores_unreliable, scoring_gaps, scoring_unverified or scoring_consistent. POST /api/evaluation-scoring-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": {
"rubric": "<rubric>"
},
"name": "wever_evaluation-scoring-check"
}
}Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| judge | No | ||
| label | No | Optional name recorded in the report. | |
| rubric | Yes | ||
| spot_check | No |