Self-evolving Context Database for AI Agents. Unify Agent Memory, Knowledge RAG and Skills.
Python agent-contextagent-memoryagent-pluginsagentic-ragcontext-databasedsh-pluginknowledge-basememorymemory-managementself-evolvingvector-database
# Copyright (c) 2026 Beijing Volcano Engine Technology Co., Ltd.
# SPDX-License-Identifier: AGPL-3.0
"""Telemetry context helpers."""
from __future__ import annotations
import contextvars
from contextlib import contextmanager
from typing import Iterator
from .operation import OperationTelemetry
_CURRENT_TELEMETRY: contextvars.ContextVar[OperationTelemetry | None] = contextvars.ContextVar(
"openviking_operation_telemetry",
default=None,
)
_CURRENT_TELEMETRY_STAGE: contextvars.ContextVar[str | None] = contextvars.ContextVar(
"openviking_operation_telemetry_stage",
default=None,
)
_CURRENT_INGESTION_SOURCE_TYPE: contextvars.ContextVar[str | None] = contextvars.ContextVar(
"openviking_ingestion_source_type",
default=None,
)
def get_current_telemetry() -> OperationTelemetry:
"""Get current operation telemetry or create a request-local disabled collector."""
telemetry = _CURRENT_TELEMETRY.get()
if telemetry is None:
telemetry = OperationTelemetry(operation="noop", enabled=False)
_CURRENT_TELEMETRY.set(telemetry)
return telemetry
@contextmanager
def bind_telemetry(handle: OperationTelemetry) -> Iterator[OperationTelemetry]:
"""Bind operation telemetry to current context."""
token = _CURRENT_TELEMETRY.set(handle)
try:
yield handle
finally:
_CURRENT_TELEMETRY.reset(token)
def get_current_telemetry_stage() -> str | None:
"""Return the currently bound low-cardinality telemetry stage, if any."""
stage = _CURRENT_TELEMETRY_STAGE.get()
if stage is None:
return None
normalized = str(stage).strip()
return normalized or None
@contextmanager
def bind_telemetry_stage(stage: str | None) -> Iterator[str | None]:
"""Bind a fixed stage label to the current context for downstream token attribution."""
normalized = None if stage is None else str(stage).strip() or None
token = _CURRENT_TELEMETRY_STAGE.set(normalized)
try:
yield normalized
finally:
_CURRENT_TELEMETRY_STAGE.reset(token)
def get_ingestion_source_type() -> str | None:
"""Return the source type bound to the current ingestion operation."""
return _CURRENT_INGESTION_SOURCE_TYPE.get()
@contextmanager
def bind_ingestion_source_type(source_type: str | None) -> Iterator[str | None]:
"""Bind resource origin for downstream model-token attribution."""
token = _CURRENT_INGESTION_SOURCE_TYPE.set(source_type)
try:
yield source_type
finally:
_CURRENT_INGESTION_SOURCE_TYPE.reset(token)
__all__ = [
"bind_ingestion_source_type",
"bind_telemetry",
"bind_telemetry_stage",
"get_current_telemetry",
"get_current_telemetry_stage",
"get_ingestion_source_type",
]