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
from types import SimpleNamespace
import pytest
from vikingbot.agent.loop import AgentLoop
from vikingbot.config.schema import AgentsConfig, SessionKey
from vikingbot.hooks.base import HookContext
from vikingbot.hooks.builtins.openviking_hooks import OpenVikingCompactHook
from vikingbot.openviking_mount.session_state import make_openviking_storage_session_id
from vikingbot.session.manager import Session
def test_session_context_defaults_to_working_memory():
assert AgentsConfig().session_context_enabled is True
@pytest.mark.asyncio
async def test_session_context_commit_uses_turn_budget_retention():
session_key = SessionKey(type="test", channel_id="channel", chat_id="chat")
session = Session(key=session_key)
session.add_message("user", "question")
session.add_message("assistant", "answer")
commit_calls: list[dict] = []
class FakeClient:
@staticmethod
def session_owner_user_id():
return None
@staticmethod
async def append_messages(**kwargs):
return {"added": len(kwargs["messages"])}
@staticmethod
async def get_session(session_id, user_id=None):
return {"session_id": session_id, "pending_tokens": 0}
@staticmethod
async def commit_session(**kwargs):
commit_calls.append(kwargs)
return {"archived": True, "effective_enable_working_memory": True}
result = await OpenVikingCompactHook()._execute_session_context_commit(
HookContext(event_type="message.compact", session_key=session_key),
session,
FakeClient(),
SimpleNamespace(commit_token_threshold=200_000),
"admin",
force_commit=True,
keep_recent_turn_count=3,
retained_message_token_budget=12_000,
min_raw_tail_steps=1,
commit_message_threshold=None,
)
assert result["success"] is True
assert commit_calls == [
{
"session_id": make_openviking_storage_session_id(session_key.safe_name()),
"keep_recent_count": 0,
"enable_working_memory": True,
"retention_mode": "turn_budget",
"keep_recent_turn_count": 3,
"retained_message_token_budget": 12_000,
"min_raw_tail_steps": 1,
"user_id": None,
}
]
def test_prompt_history_consumes_overview_checkpoint_and_raw_tail_in_order():
session_key = SessionKey(type="test", channel_id="channel", chat_id="chat")
session = Session(key=session_key)
context_payload = {
"latest_archive_overview": "Earlier turns were compacted.",
"messages": [
{
"role": "user",
"message_kind": "user_query",
"parts": [{"type": "text", "text": "Investigate the outage"}],
},
{
"role": "assistant",
"message_kind": "checkpoint",
"parts": [
{
"type": "context",
"context_type": "memory",
"uri": "viking://user/default/sessions/s/history/archive_001",
"abstract": "Checked early signals and confirmed pool saturation.",
}
],
},
{
"role": "assistant",
"message_kind": "assistant_step",
"parts": [
{"type": "text", "text": "I will verify the recovery setting."},
{
"type": "tool",
"tool_id": "call-2",
"tool_name": "read_config",
"tool_output": "recovery_timeout=30",
},
],
},
],
}
loop = AgentLoop.__new__(AgentLoop)
history = loop._build_ov_history_messages(session, context_payload)
assert [message["role"] for message in history] == [
"assistant",
"user",
"assistant",
"assistant",
]
assert history[0]["content"] == (
"[Earlier conversation summary]\nEarlier turns were compacted."
)
assert history[1]["content"] == "Investigate the outage"
assert history[2]["content"] == ("Checked early signals and confirmed pool saturation.")
assert history[3]["content"] == ("I will verify the recovery setting.\nrecovery_timeout=30")