Recall

Recall is how you fetch relevant memory context for a given task.

How Recall Works

Neocortex ranks candidate memories using:

• semantic relevance to the query

• recency/decay behavior

• interaction reinforcement signals

The output can include both structured chunks and an LLM-ready context string.

Prompt-Driven Retrieval

Most applications recall with a query such as:

• "What dietary restrictions does the user have?"

• "Summarize prior project decisions"

This keeps retrieval dynamic and task-specific.

Scope With Namespaces

Always scope recall to the namespace most relevant to the current task. This reduces noise and improves answer quality.

For implementation examples in all supported languages, see Recalling Memories.