cayu.evals.workflow_target

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.

`WORKFLOW_EVAL_DEFAULT_CLOSE_TIMEOUT_SECONDS`

from cayu.evals import WORKFLOW_EVAL_DEFAULT_CLOSE_TIMEOUT_SECONDS

WORKFLOW_EVAL_DEFAULT_CLOSE_TIMEOUT_SECONDS = 30.0

Release source · Runtime guarantees · Capability guides

`workflow_eval_input_messages_sha256`

from cayu.evals import workflow_eval_input_messages_sha256

Bind the exact candidate input retained on a workflow-root trajectory.

workflow_eval_input_messages_sha256(messages: tuple[Message, ...]) -> str

Release source · Runtime guarantees · Capability guides

`WORKFLOW_EVAL_MAX_APPLICATION_CONTEXT_BYTES`

from cayu.evals import WORKFLOW_EVAL_MAX_APPLICATION_CONTEXT_BYTES

WORKFLOW_EVAL_MAX_APPLICATION_CONTEXT_BYTES = 64 << 10

Release source · Runtime guarantees · Capability guides

`WORKFLOW_EVAL_MAX_CLOSE_TIMEOUT_SECONDS`

from cayu.evals import WORKFLOW_EVAL_MAX_CLOSE_TIMEOUT_SECONDS

WORKFLOW_EVAL_MAX_CLOSE_TIMEOUT_SECONDS = 300.0

Release source · Runtime guarantees · Capability guides

`WORKFLOW_EVAL_MAX_FINAL_OUTPUT_CHARS`

from cayu.evals import WORKFLOW_EVAL_MAX_FINAL_OUTPUT_CHARS

WORKFLOW_EVAL_MAX_FINAL_OUTPUT_CHARS = 1048576

Release source · Runtime guarantees · Capability guides

`WORKFLOW_EVAL_MAX_INPUT_BYTES`

from cayu.evals import WORKFLOW_EVAL_MAX_INPUT_BYTES

WORKFLOW_EVAL_MAX_INPUT_BYTES = 8 << 20

Release source · Runtime guarantees · Capability guides

`WORKFLOW_EVAL_MAX_INPUT_MESSAGES`

from cayu.evals import WORKFLOW_EVAL_MAX_INPUT_MESSAGES

WORKFLOW_EVAL_MAX_INPUT_MESSAGES = 32

Release source · Runtime guarantees · Capability guides

`WORKFLOW_EVAL_MAX_STRUCTURED_OUTPUT_BYTES`

from cayu.evals import WORKFLOW_EVAL_MAX_STRUCTURED_OUTPUT_BYTES

WORKFLOW_EVAL_MAX_STRUCTURED_OUTPUT_BYTES = 256 << 10

Release source · Runtime guarantees · Capability guides

`workflow_eval_output_sha256`

from cayu.evals import workflow_eval_output_sha256

Bind one projected text result without duplicating it in evidence metadata.

workflow_eval_output_sha256(final_output: str) -> str

Release source · Runtime guarantees · Capability guides

`workflow_eval_trial_session_id`

from cayu.evals import workflow_eval_trial_session_id

Derive the recovery-stable root identity for one concrete workflow trial.

workflow_eval_trial_session_id(*, target_revision: str, run_id: str, suite_id: str, case_id: str, trial_number: int) -> str

Release source · Runtime guarantees · Capability guides

`workflow_spec_revision`

from cayu.evals import workflow_spec_revision

Return the canonical public revision of one workflow specification.

workflow_spec_revision(spec: WorkflowSpec) -> str

Release source · Runtime guarantees · Capability guides

`WorkflowEvalExecution`

from cayu.evals import WorkflowEvalExecution

The exact application/workflow pair and optional quiescence callback for one trial.

Decorators: @dataclass(frozen=True, slots=True).

Fields and attributes

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
`app`	`CayuApp`	`required`
`workflow`	`WorkflowBase`	`required`
`close`	`Callable[[], Awaitable[None]] \| None`	`None`

Release source · Runtime guarantees · Capability guides

`WorkflowEvalFactory`

from cayu.evals import WorkflowEvalFactory

WorkflowEvalFactory: TypeAlias = Callable[[WorkflowEvalInvocation], WorkflowEvalExecution | Awaitable[WorkflowEvalExecution]]

Release source · Runtime guarantees · Capability guides

`WorkflowEvalInstanceScope`

from cayu.evals import WorkflowEvalInstanceScope

How a target constructs application/workflow state for concrete trials.

Bases: StrEnum.

Fields and attributes

Name	Type	Default
`SHARED`	`class attribute`	`'shared'`
`PER_TRIAL`	`class attribute`	`'per_trial'`

Release source · Runtime guarantees · Capability guides

`WorkflowEvalInvocation`

from cayu.evals import WorkflowEvalInvocation

Bounded candidate input and runtime identities supplied to one trusted factory.

Bases: BaseModel.

Fields and attributes

Name	Type	Default
`run_id`	`StrictStr`	`Field(max_length=256)`
`suite_id`	`StrictStr`	`Field(max_length=128)`
`case_id`	`StrictStr`	`Field(max_length=128)`
`trial_number`	`StrictInt`	`Field(ge=1, le=100)`
`workflow_run_id`	`StrictStr`	`Field(max_length=128)`
`idempotency_key`	`StrictStr`	`Field(max_length=256)`
`messages`	`tuple[Message, ...]`	`required`
`application_context`	`dict[str, Any]`	`Field(default_factory=dict)`

validate_ids (method)

WorkflowEvalInvocation.validate_ids(value: str, info) -> str

copy_messages (method)

WorkflowEvalInvocation.copy_messages(value: object) -> tuple[Message, ...]

copy_application_context (method)

WorkflowEvalInvocation.copy_application_context(value: object) -> dict[str, Any]

Release source · Runtime guarantees · Capability guides

`WorkflowEvalOutputEvidenceV1`

from cayu.evals import WorkflowEvalOutputEvidenceV1

Replay-safe binding from projected output to one exact workflow completion.

Bases: BaseModel.

Fields and attributes

Name	Type	Default
`schema_version`	`Literal[1]`	`1`
`target_revision`	`StrictStr`	`required`
`projector_revision`	`StrictStr`	`required`
`workflow_name`	`StrictStr`	`required`
`attempt_id`	`StrictStr`	`required`
`completion_event_id`	`StrictStr`	`required`
`input_message_count`	`StrictInt`	`Field(ge=0, le=WORKFLOW_EVAL_MAX_INPUT_MESSAGES)`
`input_messages_sha256`	`StrictStr`	`required`
`final_output_sha256`	`StrictStr`	`required`
`structured_output`	`dict[str, Any] \| None`	`None`

validate_schema_version_type (method)

WorkflowEvalOutputEvidenceV1.validate_schema_version_type(value: object) -> object

validate_revisions (method)

WorkflowEvalOutputEvidenceV1.validate_revisions(value: str, info) -> str

validate_sha256 (method)

WorkflowEvalOutputEvidenceV1.validate_sha256(value: str, info) -> str

validate_ids (method)

WorkflowEvalOutputEvidenceV1.validate_ids(value: str, info) -> str

copy_structured_output (method)

WorkflowEvalOutputEvidenceV1.copy_structured_output(value: object) -> dict[str, Any] | None

Release source · Runtime guarantees · Capability guides

`WorkflowEvalResult`

from cayu.evals import WorkflowEvalResult

Typed bounded candidate output returned by an application-owned projector.

Bases: BaseModel.

Fields and attributes

Name	Type	Default
`final_output`	`StrictStr`	`required`
`structured_output`	`dict[str, Any] \| None`	`None`

validate_final_output (method)

WorkflowEvalResult.validate_final_output(value: str, info) -> str

copy_structured_output (method)

WorkflowEvalResult.copy_structured_output(value: object) -> dict[str, Any] | None

Release source · Runtime guarantees · Capability guides

`WorkflowEvalResultProjector`

from cayu.evals import WorkflowEvalResultProjector

WorkflowEvalResultProjector: TypeAlias = Callable[[WorkflowEvalTerminalEvidence], WorkflowEvalResult | Awaitable[WorkflowEvalResult]]

Release source · Runtime guarantees · Capability guides

`WorkflowEvalTargetIdentityV1`

from cayu.evals import WorkflowEvalTargetIdentityV1

Portable behavior identity for a trusted workflow-root eval target.

Bases: BaseModel.

Fields and attributes

Name	Type	Default
`schema_version`	`Literal[1]`	`1`
`revision`	`StrictStr`	`required`
`workflow_name`	`StrictStr`	`required`
`workflow_spec_revision`	`StrictStr`	`required`
`implementation_revision`	`StrictStr`	`required`
`result_projector_revision`	`StrictStr`	`required`
`execution_scope_revision`	`StrictStr`	`required`
`application_context_revision`	`StrictStr`	`required`
`evidence_policy_revision`	`StrictStr`	`required`
`capture_bounds`	`SessionTrajectoryBounds \| None`	`Field(default=None, exclude_if=lambda value: value is None)`
`instance_scope`	`WorkflowEvalInstanceScope`	`required`
`close_timeout_seconds`	`float`	`Field(gt=0, le=WORKFLOW_EVAL_MAX_CLOSE_TIMEOUT_SECONDS)`

validate_schema_version_type (method)

WorkflowEvalTargetIdentityV1.validate_schema_version_type(value: object) -> object

validate_workflow_name (method)

WorkflowEvalTargetIdentityV1.validate_workflow_name(value: str, info) -> str

validate_revisions (method)

WorkflowEvalTargetIdentityV1.validate_revisions(value: str, info) -> str

validate_close_timeout (method)

WorkflowEvalTargetIdentityV1.validate_close_timeout(value: object) -> object

validate_revision (method)

WorkflowEvalTargetIdentityV1.validate_revision() -> WorkflowEvalTargetIdentityV1

create (method)

WorkflowEvalTargetIdentityV1.create(*, workflow_spec: WorkflowSpec, implementation_revision: str, result_projector_revision: str, execution_scope_revision: str, application_context: dict[str, Any], evidence_policy_revision: str, instance_scope: WorkflowEvalInstanceScope | str, close_timeout_seconds: float, capture_bounds: SessionTrajectoryBounds | None=None) -> WorkflowEvalTargetIdentityV1

Release source · Runtime guarantees · Capability guides

`WorkflowEvalTerminalEvidence`

from cayu.evals import WorkflowEvalTerminalEvidence

Exact current-attempt completion evidence visible to the result projector.

Bases: BaseModel.

Fields and attributes

Name	Type	Default
`workflow_run_id`	`StrictStr`	`required`
`workflow_name`	`StrictStr`	`required`
`attempt_id`	`StrictStr`	`required`
`completion_event`	`Event`	`required`

validate_ids (method)

WorkflowEvalTerminalEvidence.validate_ids(value: str, info) -> str

copy_completion_event (method)

WorkflowEvalTerminalEvidence.copy_completion_event(value: object) -> object

validate_completion_boundary (method)

WorkflowEvalTerminalEvidence.validate_completion_boundary() -> WorkflowEvalTerminalEvidence

Release source · Runtime guarantees · Capability guides