How Memory Recall Works
Why an agent brings back certain memories in a conversation and not others.
Memory recall is based on meaning, not on a chronological log. Knowing this helps you predict which memories will actually surface in a given conversation.
How a Memory Becomes Searchable
When a memory is saved — whether you added it from the Memory tab or the agent saved it on its own — its name and content are automatically turned into a form the agent can search by meaning, rather than by matching exact words. If you or the agent later edit a memory's name or content, it's automatically reprocessed so recall keeps matching the updated text.
What's Brought Back During a Chat
When you chat with an agent, it isn't simply shown its most recently added memories. Instead, it's shown the memories whose meaning is most relevant to what you're currently discussing. That's why a memory saved weeks ago can resurface at exactly the right moment, and why unrelated memories usually stay out of the conversation even if they were added more recently.