Overview

The Key-value store node provides project-scoped persistent storage for any Data Value type, enabling stateful workflows and cross-flow data sharing. This node serves as a flexible storage mechanism that supports Insert, Retrieve, and Remove operations with sophisticated scoping and isolation capabilities.

The node supports advanced key name-spacing through special prefixes ($flow., $session., $run.) that scope data to different execution contexts, and optional entity isolation for multi-tenant or per-user data separation. It can store and retrieve any supported Data Value type including Text, Numbers, JSON objects, Content, Segments, Embeddings, Knowledge, Knowledge DBs, LLM Models, and Configured services. This makes it essential for maintaining application state, implementing per-user memory systems, building conversational context, and enabling data persistence across flow executions.

Node Specification

Core Properties

Property Value
Node Type Persistent storage
Category Data flow

Parameters

Parameter Type Options Default Required Description
Operation Choice Insert, Retrieve, Remove Insert Yes Storage operation to perform
Data type Choice Text, Numbers, JSON object, Content, Segments, Embeddings, Knowledge, Knowledge DB, LLM Model, Configured service Text Yes Type of value to store/retrieve (configures input/output ports)
Key Text (empty) No Storage key; supports $flow., $session., $run. prefixes for scoping
Memory store Entity ID Text (empty) No Optional entity identifier for per-entity isolation (e.g., user ID, email)

Input Ports

Port Type Accepts Required Description
Key Text Text No Storage key (can be wired dynamically; supports scoping prefixes)
Value to store Configurable Based on Data type No Value to insert (used when Operation = Insert)
Store Entity ID Text Text No Optional entity isolation key

Output Ports

Port Type Description
Retrieved value Configurable Value retrieved when Operation = Retrieve (matches Data type)

Process Flow Exits

Exit Condition Description
Main exit Always Continues to downstream nodes

Common Flows

Basic Key-Value Operations

Data Input → Key-value Store (Insert) → Downstream Processing
           ↓
Key Input → Key-value Store (Retrieve) → Retrieved Data

Store data and retrieve it later in the same or different flows

User Memory System

User ID → Key-value Store (Entity ID + Insert) → Per-User Data Storage
       ↓
User ID → Key-value Store (Entity ID + Retrieve) → User-Specific Data