Actions46
- Bank Actions
- Directive Actions
- Document Actions
- Entity Actions
- Memory Actions
- Mental Model Actions
- Operation Actions
- Webhook Actions
Memory → Recall
AI-generatedSummary
Searches and retrieves relevant memories from a specified memory bank using a natural language query and optional filters.
Inputs
- Bank ID (required) — ID of the bank to search memories in.
- Query (required) — Natural language search query.
- Additional Fields — Optional parameters to refine the recall, including Budget for retrieval depth (Low, Mid, High), inclusion flags for raw source chunks, entity observations, and source facts, Max Tokens limit for returned facts, Query Timestamp to anchor relative temporal expressions, Tags to filter memories, Tags Match to determine how to filter multiple tags, Trace to enable debug trace data, and Types to specify memory types to include (defaults to world and experience if empty).
Output shape
a JSON array containing recalled memory records relevant to the query, filtered and detailed according to the input parameters.
The result structure depends on the retrieved memories and the included optional fields like entities, chunks, and source facts. Pagination is not described, and error responses are handled by the node's error handling.