Preview draft judge grading on one recorded iteration
Uses model budget and complete recorded evidence without changing the run verdict. Supply rubric.instructions, optional structured criteria, or null for objective-only grading. Continue with the unchanged rubric and returned cursor, sourceHash and reservationId. Completed page retries return cached results. Error cases have a null draft measurement.
Path parameters
ID of the hosted project that contains the server.
Eval run ID, as returned by POST /eval-runs.
Headers
Which vocabulary this request and its response speak. Absent means 1, which is byte-for-byte today's contract: the same request fields, the same refusals, the same response projection. 2 is the canonical vocabulary. Any other value is a 400 with code: "VALIDATION_ERROR".
Today it decides one thing: the spelling of an evaluator's policy role. Vocabulary 1 accepts and returns gating; vocabulary 2 accepts both spellings and returns the canonical required. Sending required without the header is a 400, deliberately — vocabulary 1 is not widened to meet vocabulary 2 half way, because a boundary that accepts a spelling it does not announce is one two implementations can disagree about.
A response that varies by vocabulary sends Vary: x-mcpjam-eval-vocabulary.
Request body
Response
One cached or newly graded page. Errors are unscored; isDone indicates the end of the population.