cayu.evals.knowledge_maintenance

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.

KNOWLEDGE_MAINTENANCE_EVALUATION_CORPUS_SCHEMA_VERSION

from cayu.evals import KNOWLEDGE_MAINTENANCE_EVALUATION_CORPUS_SCHEMA_VERSION

KNOWLEDGE_MAINTENANCE_EVALUATION_CORPUS_SCHEMA_VERSION = 'cayu.knowledge_maintenance_evaluation_corpus.v1'

Release source · Runtime guarantees · Capability guides

KNOWLEDGE_MAINTENANCE_EVALUATION_RESULT_SCHEMA_VERSION

from cayu.evals import KNOWLEDGE_MAINTENANCE_EVALUATION_RESULT_SCHEMA_VERSION

KNOWLEDGE_MAINTENANCE_EVALUATION_RESULT_SCHEMA_VERSION = 'cayu.knowledge_maintenance_evaluation_result.v1'

Release source · Runtime guarantees · Capability guides

KnowledgeMaintenanceEvaluationCase

from cayu.evals import KnowledgeMaintenanceEvaluationCase

Bases: BaseModel.

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
id	str	required
scenario	KnowledgeMaintenanceEvaluationScenario	required
entries	tuple[KnowledgeMaintenanceEvaluationEntry, ...]	required
source_entry_ids	tuple[str, ...]	required
signal_kind	KnowledgeMaintenanceSignalKind	required
expected_routed_entry_ids	tuple[str, ...]	required
replacement_title	str	required
replacement_text	str	required
claims	tuple[KnowledgeMaintenanceEvaluationClaim, ...]	required
dispositions	tuple[KnowledgeMaintenanceEvaluationDisposition, ...]	required
evaluator_verdict	KnowledgeMaintenanceEvaluationVerdict	required
review_decision	KnowledgeMaintenanceDecisionKind | None	None
advance_source_before_review	str | None	None
recall_query	str	required

validate_id (method)

KnowledgeMaintenanceEvaluationCase.validate_id(value: str) -> str

validate_text (method)

KnowledgeMaintenanceEvaluationCase.validate_text(value: str, info) -> str

validate_recall_query (method)

KnowledgeMaintenanceEvaluationCase.validate_recall_query(value: str) -> str

validate_references (method)

KnowledgeMaintenanceEvaluationCase.validate_references() -> KnowledgeMaintenanceEvaluationCase

relation_kinds (property)

KnowledgeMaintenanceEvaluationCase.relation_kinds() -> dict[str, KnowledgeRelationKind]

Release source · Runtime guarantees · Capability guides

KnowledgeMaintenanceEvaluationCaseResult

from cayu.evals import KnowledgeMaintenanceEvaluationCaseResult

Bases: BaseModel.

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
case_id	str	required
scenario	KnowledgeMaintenanceEvaluationScenario	required
routed_entry_ids	tuple[str, ...]	required
routing_precision	float	required
routing_recall	float	required
planning_outcome	KnowledgeMaintenancePlanningOutcome	required
storage_outcome	Literal['not_published', 'pending', 'applied', 'rejected', 'stale']	required
information_retention	float	required
evidence_retention	float | None	required
unsafe_acceptance_count	int	required
lifecycle_correct	bool	required
lineage_correct	bool	required
recalled_entry_ids	tuple[str, ...]	required
model_call_count	int	required
latency_ms	float	required

validate_rate (method)

KnowledgeMaintenanceEvaluationCaseResult.validate_rate(value: float | None, info) -> float | None

validate_unsafe_acceptance_count (method)

KnowledgeMaintenanceEvaluationCaseResult.validate_unsafe_acceptance_count(value) -> int

validate_model_call_count (method)

KnowledgeMaintenanceEvaluationCaseResult.validate_model_call_count(value) -> int

validate_latency (method)

KnowledgeMaintenanceEvaluationCaseResult.validate_latency(value: float) -> float

validate_selected_ids (method)

KnowledgeMaintenanceEvaluationCaseResult.validate_selected_ids() -> KnowledgeMaintenanceEvaluationCaseResult

Release source · Runtime guarantees · Capability guides

KnowledgeMaintenanceEvaluationClaim

from cayu.evals import KnowledgeMaintenanceEvaluationClaim

Bases: BaseModel.

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
id	str	required
text	str	required
source_entry_ids	tuple[str, ...]	required

validate_id (method)

KnowledgeMaintenanceEvaluationClaim.validate_id(value: str) -> str

validate_text (method)

KnowledgeMaintenanceEvaluationClaim.validate_text(value: str) -> str

validate_sources (method)

KnowledgeMaintenanceEvaluationClaim.validate_sources() -> KnowledgeMaintenanceEvaluationClaim

Release source · Runtime guarantees · Capability guides

KnowledgeMaintenanceEvaluationCorpus

from cayu.evals import KnowledgeMaintenanceEvaluationCorpus

Portable public or private maintenance-quality corpus.

Bases: BaseModel.

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
schema_version	Literal['cayu.knowledge_maintenance_evaluation_corpus.v1']	required
corpus_revision	str	required
origin	Literal['hermetic_public', 'external_private']	required
cases	tuple[KnowledgeMaintenanceEvaluationCase, ...]	required

validate_revision (method)

KnowledgeMaintenanceEvaluationCorpus.validate_revision(value: str) -> str

validate_cases (method)

KnowledgeMaintenanceEvaluationCorpus.validate_cases() -> KnowledgeMaintenanceEvaluationCorpus

Release source · Runtime guarantees · Capability guides

KnowledgeMaintenanceEvaluationDisposition

from cayu.evals import KnowledgeMaintenanceEvaluationDisposition

Bases: BaseModel.

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
source_entry_id	str	required
relation_kind	KnowledgeRelationKind	required

validate_source_entry_id (method)

KnowledgeMaintenanceEvaluationDisposition.validate_source_entry_id(value: str) -> str

Release source · Runtime guarantees · Capability guides

KnowledgeMaintenanceEvaluationEntry

from cayu.evals import KnowledgeMaintenanceEvaluationEntry

Bases: BaseModel.

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
id	str	required
text	str	required

validate_id (method)

KnowledgeMaintenanceEvaluationEntry.validate_id(value: str) -> str

validate_text (method)

KnowledgeMaintenanceEvaluationEntry.validate_text(value: str) -> str

Release source · Runtime guarantees · Capability guides

KnowledgeMaintenanceEvaluationMetrics

from cayu.evals import KnowledgeMaintenanceEvaluationMetrics

Bases: BaseModel.

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
case_count	int	required
routing_precision	float	required
routing_recall	float	required
information_retention	float	required
evidence_retention	float	required
unsafe_acceptance_rate	float	required
lifecycle_correctness	float	required
lineage_correctness	float	required
model_call_count	int	required
latency_p50_ms	float	required
latency_p95_ms	float	required

validate_case_count (method)

KnowledgeMaintenanceEvaluationMetrics.validate_case_count(value) -> int

validate_rate (method)

KnowledgeMaintenanceEvaluationMetrics.validate_rate(value: float, info) -> float

validate_model_call_count (method)

KnowledgeMaintenanceEvaluationMetrics.validate_model_call_count(value) -> int

validate_latency (method)

KnowledgeMaintenanceEvaluationMetrics.validate_latency(value: float, info) -> float

validate_percentiles (method)

KnowledgeMaintenanceEvaluationMetrics.validate_percentiles() -> KnowledgeMaintenanceEvaluationMetrics

Release source · Runtime guarantees · Capability guides

KnowledgeMaintenanceEvaluationResult

from cayu.evals import KnowledgeMaintenanceEvaluationResult

Bases: BaseModel.

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
schema_version	Literal['cayu.knowledge_maintenance_evaluation_result.v1']	required
corpus_revision	str	required
corpus_origin	Literal['hermetic_public', 'external_private']	required
backend	str	required
execution_mode	Literal['deterministic_fixture']	'deterministic_fixture'
planner_identity	str	required
evaluator_identity	str	required
configuration	dict[str, Any]	required
metrics	KnowledgeMaintenanceEvaluationMetrics	required
cases	tuple[KnowledgeMaintenanceEvaluationCaseResult, ...]	required

validate_identity (method)

KnowledgeMaintenanceEvaluationResult.validate_identity(value: str, info) -> str

copy_configuration (method)

KnowledgeMaintenanceEvaluationResult.copy_configuration(value) -> dict[str, Any]

validate_case_aggregate (method)

KnowledgeMaintenanceEvaluationResult.validate_case_aggregate() -> KnowledgeMaintenanceEvaluationResult

Release source · Runtime guarantees · Capability guides

KnowledgeMaintenanceEvaluationScenario

from cayu.evals import KnowledgeMaintenanceEvaluationScenario

Reference outcomes required by the reviewed-maintenance architecture.

Bases: StrEnum.

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
DUPLICATE_MERGE	class attribute	'duplicate_merge'
AUTHORITATIVE_SUPERSESSION	class attribute	'authoritative_supersession'
UNRESOLVED_CONTRADICTION	class attribute	'unresolved_contradiction'
STALE_PROPOSAL	class attribute	'stale_proposal'
REVIEWER_REJECTION	class attribute	'reviewer_rejection'
HISTORICAL_LINEAGE	class attribute	'historical_lineage'

Release source · Runtime guarantees · Capability guides

load_knowledge_maintenance_evaluation_corpus

from cayu.evals import load_knowledge_maintenance_evaluation_corpus

load_knowledge_maintenance_evaluation_corpus(path: str | Path) -> KnowledgeMaintenanceEvaluationCorpus

Release source · Runtime guarantees · Capability guides

run_knowledge_maintenance_evaluation

from cayu.evals import run_knowledge_maintenance_evaluation

Run exact maintenance workflows against an empty knowledge store.

The injected components are deliberately deterministic and have zero model-call authority. This evaluates Cayu’s orchestration and safety contracts separately from provider-dependent planner/evaluator quality.

async run_knowledge_maintenance_evaluation(corpus: KnowledgeMaintenanceEvaluationCorpus, store: KnowledgeStore, *, backend: str, configuration: dict[str, Any] | None=None) -> KnowledgeMaintenanceEvaluationResult

Release source · Runtime guarantees · Capability guides