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
from types import SimpleNamespace
import pytest
from openviking.privacy import scan as privacy_scan
from openviking.privacy.sanitizer import PrivacyReplacement, PrivacySanitizationResult
from openviking.server.identity import RequestContext, Role
from openviking_cli.exceptions import InvalidArgumentError, ResourceExhaustedError
from openviking_cli.session.user_id import UserIdentifier
def _ctx() -> RequestContext:
return RequestContext(user=UserIdentifier("account-a", "alice"), role=Role.USER)
class _VikingFS:
def __init__(self):
self.read_limits: list[int | None] = []
async def stat(self, uri, *, ctx, skip_count): # noqa: ANN001
del ctx, skip_count
if uri == "viking://resources/demo":
return {"isDir": True, "size": 0}
raise AssertionError(f"unexpected stat: {uri}")
async def tree(self, uri, **kwargs): # noqa: ANN001
assert uri == "viking://resources/demo"
assert kwargs["node_limit"] == privacy_scan.PRIVACY_SCAN_MAX_NODES
return [
{
"uri": "viking://resources/demo/large.md",
"size": privacy_scan.PRIVACY_SCAN_MAX_FILE_BYTES + 1,
"isDir": False,
},
{
"uri": "viking://resources/demo/content.md",
"size": 13,
"isDir": False,
},
{
"uri": "viking://resources/demo/.abstract.md",
"size": 10,
"isDir": False,
},
]
async def read_file_bytes(self, uri, *, ctx, max_bytes): # noqa: ANN001
del ctx
assert uri == "viking://resources/demo/content.md"
self.read_limits.append(max_bytes)
return b"prefix secret"
class _Sanitizer:
async def sanitize( # noqa: ANN001
self,
content,
*,
ctx,
uri,
max_replacements,
fail_on_match_limit,
):
del ctx
assert content == "prefix secret"
assert uri == "viking://resources/demo/content.md"
assert max_replacements == privacy_scan.PRIVACY_SCAN_MAX_REPLACEMENTS_PER_FILE
assert fail_on_match_limit is False
return PrivacySanitizationResult(
content="prefix reference",
replacements=(PrivacyReplacement(7, 13, "reference"),),
)
async def test_privacy_scan_applies_file_and_node_budgets_before_reading():
viking_fs = _VikingFS()
service = SimpleNamespace(viking_fs=viking_fs)
result = await privacy_scan.scan_privacy_content(
service=service,
ctx=_ctx(),
sanitizer=_Sanitizer(),
uri="viking://resources/demo",
)
assert result["enumerated_files"] == 3
assert result["scanned_files"] == 1
assert result["matched_files"] == 1
assert result["skipped_files"] == {"file_too_large": 1, "internal_metadata": 1}
assert viking_fs.read_limits == [privacy_scan.PRIVACY_SCAN_MAX_FILE_BYTES]
async def test_privacy_scan_rejects_invalid_budget_and_concurrent_run():
service = SimpleNamespace(viking_fs=_VikingFS())
with pytest.raises(InvalidArgumentError, match="node_limit"):
await privacy_scan.scan_privacy_content(
service=service,
ctx=_ctx(),
sanitizer=_Sanitizer(),
uri="viking://resources/demo",
node_limit=privacy_scan.PRIVACY_SCAN_MAX_NODES + 1,
)
privacy_scan._SCAN_ACTIVE = True
try:
with pytest.raises(ResourceExhaustedError, match="already running"):
await privacy_scan.scan_privacy_content(
service=service,
ctx=_ctx(),
sanitizer=_Sanitizer(),
uri="viking://resources/demo",
)
finally:
privacy_scan._SCAN_ACTIVE = False