List design eval runs

GET/v1/design/evals/runs

Signed-in sessionLaunch key product.design.evalsOperation design_eval_runs_list

An internal tool for design rules and components that agents synthesize, review and promote, with evaluation runs to check them.

Query parameters

expectedAuthorizationVersionstring
split"challenge" | "golden" | "holdout"Required
candidateBundleIdstringRequired
cursorstring | null
limitinteger

Returns

itemsobject[]Required
nextCursorstring | nullRequired

Errors

400, 401, 403, 404, 409, 429 and 503 return {"error": {"code", "message", "correlationId"}}.

GET/v1/design/evals/runs

curl "$SPECIFY_API_ORIGIN/v1/design/evals/runs?split=<split>&candidateBundleId=<candidateBundleId>" \
  -H "Authorization: Bearer $SPECIFY_SESSION_TOKEN" \
  -H "X-Specify-Account-Id: <accountId>" \
  -H "X-Specify-Service-Tenancy-Id: <serviceTenancyId>"

Response

{
  "result": {
    "items": [
      {
        "runId": "<runId>",
        "createdAt": 1,
        "updatedAt": 1,
        "submittedAt": "<submittedAt>",
        "purpose": "observational",
        "split": "challenge",
        "status": "failed",
        "candidateHash": "<candidateHash>",
        "baselineHash": "<baselineHash>",
        "model": {
          "provider": "<provider>",
          "name": "<name>",
          "version": "<version>",
          "settingsHash": "<settingsHash>"
        },
        "corpusVersion": "<corpusVersion>",
        "receiptDigest": "<receiptDigest>",
        "valid": true,
        "targetDelta": "<targetDelta>",
        "hardChecksPassed": false,
        "gate": "<gate>",
        "unmetReasons": [
          "category_regression"
        ],
        "categories": [
          {
            "key": "<key>",
            "passed": true,
            "delta": 1
          }
        ],
        "observational": true,
        "promotionEligible": false
      }
    ],
    "nextCursor": "<nextCursor>"
  }
}

List design eval runs

Headers

Query parameters

Returns

Errors