Module eval 

Evaluations. The judge-backed kinds pose a criterion in plain English and ask the provider’s judge to score the transcript (a boolean assertion, or a numeric score compared against a threshold). The deterministic kinds (called / not_called) assert on the mock/spy channel’s observed tool calls — no judge, no flakiness — referencing a mocks: declaration by name.

Structs

BooleanEval

Assert a plain-English criterion holds (or, with expected: false, that it does not).

CalledEval

Deterministic: assert the referenced mock/spy observed at least one matching call (or exactly times). Evaluated against the mock channel’s records, not by a judge.

EvalOutcome

The result of running one eval against a transcript.

NotCalledEval

Deterministic: assert the referenced mock/spy observed no matching call.

NumericEval

Score a plain-English criterion on a numeric scale and compare it to a threshold.

Enums

Comparator

How a numeric score is compared to its threshold.

Eval

An eval specification, as written in a test case’s YAML (or compiled by an SDK’s case builders).

EvalDetail

The kind-specific detail of an eval outcome, for reporting.

JudgeValue

The raw value a judge returns: either a boolean or a number, matching the eval kind. Deserialized untagged from the provider’s value field.