Attributes

Configuration options for the Memory class

Attributes

The Memory system is configured through the Memory class, which provides the following attributes:

Attribute	Type	Default	Description
storage	Storage	(required)	Storage backend for memory persistence
session_id	str | None	None	Session identifier for conversation/summary memory
user_id	str | None	None	User identifier for user analysis memory
full_session_memory	bool	True	Enable complete chat history
summary_memory	bool	False	Enable conversation summaries
user_analysis_memory	bool	False	Enable user learning and profiles
num_last_messages	int | None	None	Limit history to last N message turns
model	str | Model | None	None	Model for analysis and summaries
user_profile_schema	BaseModel | None	None	Custom user profile schema
dynamic_user_profile	bool	True	Allow agent to create custom profile fields
user_memory_mode	str	"update"	User profile update mode: ‘update’ or ‘replace’
feed_tool_call_results	bool	False	Include tool calls in memory
debug	bool	False	Enable debug logging

Configuration Example

from upsonic import Agent, Task, Memory
from upsonic.storage.providers.sqlite import SqliteStorage

# Create storage
storage = SqliteStorage("sessions", "profiles", "agent_memory.db")

# Create memory with configuration
memory = Memory(
    storage=storage,
    session_id="session_001",
    user_id="user_123",
    full_session_memory=True,
    summary_memory=True,
    user_analysis_memory=True,
    num_last_messages=20,
    model="openai/gpt-4o-mini",
    feed_tool_call_results=False,
    debug=True
)

# Create agent with memory
agent = Agent("openai/gpt-4o", memory=memory)

# Use agent - memory is automatic
task = Task("Hello! I'm interested in learning Python")
result = agent.do(task)
print(result)