Skip to content

cayu.evals.result_contract

Public Cayu API. Definitions are grouped by their public source module, and definitions from private implementation modules appear on their package’s page; adding a symbol does not move existing links. The import shown at each entry is a supported public namespace. Read the Python API overview and install the required extra.

from cayu.evals import EVAL_TRIAL_OUTPUT_MAX_PREVIEW_BYTES
EVAL_TRIAL_OUTPUT_MAX_PREVIEW_BYTES = 16 << 10

Release source · Runtime guarantees · Capability guides

from cayu.evals import EVAL_TRIAL_OUTPUT_MAX_RETAINED_BYTES
EVAL_TRIAL_OUTPUT_MAX_RETAINED_BYTES = EVAL_TRIAL_OUTPUT_MAX_RETAINED_CHARS * 4

Release source · Runtime guarantees · Capability guides

from cayu.evals import EVAL_TRIAL_OUTPUT_MAX_RETAINED_CHARS
EVAL_TRIAL_OUTPUT_MAX_RETAINED_CHARS = 65536

Release source · Runtime guarantees · Capability guides

from cayu.evals import EvalTrialDiagnosticCode

Stable, non-secret reason for one fresh trial’s terminal outcome.

Bases: StrEnum.

Defaults below are source expressions; Field(...) includes schema/validation settings and default_factory creates a fresh value. Inherited declarations are included. Constructor behavior follows the model or dataclass shown in the bases/decorators.

Name Type Default
PASSED class attribute 'passed'
RECOVERY_REEXECUTION_BLOCKED class attribute 'recovery_reexecution_blocked'
ASSERTION_FAILED class attribute 'assertion_failed'
ASSERTION_EVIDENCE_UNAVAILABLE class attribute 'assertion_evidence_unavailable'
TERMINAL_EVIDENCE_UNAVAILABLE class attribute 'terminal_evidence_unavailable'
INTERRUPTED_EVIDENCE_UNAVAILABLE class attribute 'interrupted_evidence_unavailable'
CHILD_EVIDENCE_UNAVAILABLE class attribute 'child_evidence_unavailable'
EXTERNAL_TARGET_UNAVAILABLE class attribute 'external_target_unavailable'
EXTERNAL_TARGET_CANCELLED class attribute 'external_target_cancelled'
EXTERNAL_TARGET_UNKNOWN class attribute 'external_target_unknown'
EXTERNAL_TARGET_INCOMPLETE class attribute 'external_target_incomplete'
EXTERNAL_TARGET_IDENTITY_MISMATCH class attribute 'external_target_identity_mismatch'
EXTERNAL_TARGET_FAILED class attribute 'external_target_failed'
WORKFLOW_TARGET_FAILED class attribute 'workflow_target_failed'
WORKFLOW_EXECUTION_FAILED class attribute 'workflow_execution_failed'
WORKFLOW_COMPLETION_MISSING class attribute 'workflow_completion_missing'
WORKFLOW_COMPLETION_CONFLICT class attribute 'workflow_completion_conflict'
WORKFLOW_ATTEMPT_SUPERSEDED class attribute 'workflow_attempt_superseded'
WORKFLOW_PROJECTOR_FAILED class attribute 'workflow_projector_failed'
WORKFLOW_OUTPUT_INVALID class attribute 'workflow_output_invalid'
WORKFLOW_CAPTURE_FAILED class attribute 'workflow_capture_failed'
WORKFLOW_QUIESCENCE_FAILED class attribute 'workflow_quiescence_failed'
EXECUTION_FAILED class attribute 'execution_failed'
SESSION_FAILED class attribute 'session_failed'
TERMINAL_EVIDENCE_FAILED class attribute 'terminal_evidence_failed'
EVIDENCE_PREPARATION_FAILED class attribute 'evidence_preparation_failed'
ASSERTION_EVALUATION_FAILED class attribute 'assertion_evaluation_failed'
CASE_TIMEOUT class attribute 'case_timeout'

Release source · Runtime guarantees · Capability guides

from cayu.evals import EvalTrialOutputPreviewV1

Bounded redacted output evidence retained for safe result inspection.

Bases: BaseModel.

Defaults below are source expressions; Field(...) includes schema/validation settings and default_factory creates a fresh value. Inherited declarations are included. Constructor behavior follows the model or dataclass shown in the bases/decorators.

Name Type Default
schema_version Literal[1] 1
text StrictStr ''
evidence_state EvalTrialOutputEvidenceState required
preview_truncated StrictBool False
retained_chars StrictInt Field(ge=0, le=EVAL_TRIAL_OUTPUT_MAX_RETAINED_CHARS)
retained_bytes StrictInt Field(ge=0, le=EVAL_TRIAL_OUTPUT_MAX_RETAINED_BYTES)
retained_sha256 StrictStr | None None
EvalTrialOutputPreviewV1.validate_schema_version_type(value: object) -> object
EvalTrialOutputPreviewV1.validate_text(value: str) -> str
EvalTrialOutputPreviewV1.validate_digest(value: str | None) -> str | None
EvalTrialOutputPreviewV1.validate_contract() -> EvalTrialOutputPreviewV1
EvalTrialOutputPreviewV1.unavailable() -> EvalTrialOutputPreviewV1
EvalTrialOutputPreviewV1.from_retained_evidence(text: str, evidence_state: EvalTrialOutputEvidenceState, *, max_preview_bytes: int) -> EvalTrialOutputPreviewV1

Release source · Runtime guarantees · Capability guides

PUBLISHED_EVAL_OUTPUT_PREVIEW_BUDGET_BYTES

Section titled “PUBLISHED_EVAL_OUTPUT_PREVIEW_BUDGET_BYTES”
from cayu.evals import PUBLISHED_EVAL_OUTPUT_PREVIEW_BUDGET_BYTES
PUBLISHED_EVAL_OUTPUT_PREVIEW_BUDGET_BYTES = 2 << 20

Release source · Runtime guarantees · Capability guides