Skip to main content
Sessions are lightweight temporal groupings. A memory has one primary session_id, and sequence_index preserves its order within that session.

​
Create a session

from mneno import MemoryClient

client = MemoryClient()
session = client.create_session(title="Mneno alpha release")

client.add("Prepared the release workflow.", session_id=session.id)
client.add("Published the alpha to PyPI.", session_id=session.id)
New sessions become active by default. Search and context building can boost memories from the active session and explain that boost.

​
Inspect session memories

memories = client.list_session_memories(session.id)

for memory in memories:
    print(memory.sequence_index, memory.content)
You can also attach an existing memory with client.add_memory_to_session(memory_id, session.id).

​
Close or archive sessions

closed = client.close_session(session.id)
print(closed.summary)

archived = client.archive_session(session.id)
print(archived.status)
Closing or archiving a session never deletes its memories. If you do not provide a close summary, Mneno creates one deterministically.

​
Build a timeline

timeline = client.build_timeline(session_ids=[session.id])

for event in timeline.events:
    print(event.timestamp, event.sequence_index, event.content)
Timeline ordering is deterministic: timestamp, session ID, sequence index, then memory ID.

​
Recover continuity

related = client.find_related_sessions("PyPI release")
print(related.continuity_summary)

context = client.get_session_context(
    session.id,
    "What happened during the release?",
)
Continuity uses local relevance signals. Archived sessions are not treated as active continuity sources by default.