Memory Recall

Memory recall is the retrieval step in agent memory: surfacing relevant past memory into the current context window so the model can use it. Recall is what turns stored memory into useful memory — without it, the agent has data it cannot reach. Recall mechanisms include semantic search over embeddings, structured queries (by user, by date, by tag), graph traversal, and explicit tool calls the agent invokes during its loop. Recall quality is bounded by the storage representation (vector vs structured vs graph) and the retrieval algorithm; mismatched recall is the most common cause of agents that "should remember" but do not.