Skip to content

cayu.tools.named_checks

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.tools import NamedCheck

Immutable application-owned declaration for one bounded process check.

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

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
name str required
description str required
timeout_s int required
max_output_bytes int required
execution_profile_identity ExecutionProfileBehaviorIdentity required
required_executables tuple[str, ...] required
NamedCheck.__init__(*, name: str, description: str, command: ExecCommand, timeout_s: int=DEFAULT_TIMEOUT_SECONDS, max_output_bytes: int=DEFAULT_OUTPUT_LIMIT_BYTES, execution_profile_identity: ExecutionProfileBehaviorIdentity, required_executables: Iterable[str]=()) -> None
NamedCheck.command() -> ExecCommand

Return a detached command so callers cannot mutate the declaration.

NamedCheck.profile_fingerprint() -> str

Return the stable secret-free fingerprint for this declaration.

Release source · Runtime guarantees · Capability guides

from cayu.tools import RunCheckTool

Run one application-owned check selected by its finite public name.

Bases: Tool, DurableToolRecovery.

RunCheckTool.__init__(*, checks: Iterable[NamedCheck], command_policy: CommandPolicy, max_model_output_bytes: int=DEFAULT_CHECK_MODEL_OUTPUT_BYTES, toolchain_profile: DockerCodingToolchainProfile | None=None, workspace_observation_limits: WorkspaceRevisionObservationLimits | None=None) -> None
RunCheckTool.name() -> str
RunCheckTool.execution_profile_identity() -> ExecutionProfileBehaviorIdentity | None

Return the stable behavior declaration frozen with this tool’s spec.

RunCheckTool.retain_arguments_for_model() -> bool

Retain bounded redacted original inputs privately for the same profile.

Independent of audit publication; oversized and compacted-away inputs may be omitted. This does not grant knowledge read or activation rights.

RunCheckTool.description() -> str
RunCheckTool.schema() -> dict[str, Any]
async RunCheckTool.run(ctx: ToolContext, args: dict) -> ToolResult
async RunCheckTool.reconcile_durable_tool_call(*, parent_session_id: str, parent_run_epoch: int, execution_profile_fingerprint: str | None, environment_name: str | None, environment_allocation_fingerprint: str | None, model_step_id: str, model_attempt_id: str, tool_round_id: str, tool_call_id: str, idempotency_key: str, arguments: dict[str, Any], started: bool, load_operation: Callable[[str], Awaitable[dict[str, Any] | None]], recovery_authority: DurableToolRecoveryAuthority | None=None) -> DurableToolRecoveryEvidence | None
async RunCheckTool.inspect_durable_tool_call(*, identity: DurableToolRecoveryIdentity, load_operation, observe_receipt) -> bool
RunCheckTool.checks() -> tuple[NamedCheck, ...]

Return immutable named declarations in canonical name order.

RunCheckTool.command_policy() -> CommandPolicy

Return the mandatory defense-in-depth command policy.

RunCheckTool.toolchain_profile() -> DockerCodingToolchainProfile | None

Release source · Runtime guarantees · Capability guides