Skip to content

Installation

Requirements

Dependency Supported versions
n8n 2.42.6 tested by the end-to-end suite, and nightly on the latest release; releases that ship @n8n/ai-node-sdk
@n8n/ai-node-sdk 0.32.4 or later, provided by n8n
Redis Agent Memory a store on an Agent Memory service, with an API key

Older n8n versions that ship @n8n/ai-node-sdk may work but are not tested.

Install the node

  1. Open Settings → Community Nodes.
  2. Choose Install, enter @redis/n8n-nodes-agent-memory, accept the risk notice and install.
  3. Three entries appear in the node picker: Redis Agent Chat Memory under AI → Memory, Redis Agent Memory in the main node list, and Redis Agent Memory Tool in the AI Agent's Tool picker.

This is the path for self-hosted n8n and for n8n Cloud once the node is verified. See the n8n community nodes guide for the current rules.

In the n8n installation directory (for Docker, inside the container's ~/.n8n/nodes folder):

npm install @redis/n8n-nodes-agent-memory

Restart n8n afterwards.

git clone https://github.com/redis/n8n-nodes-agent-memory.git
cd n8n-nodes-agent-memory
npm install
npm run dev        # builds and starts n8n with this node at http://localhost:5678

npm run dev runs n8n in a container from n8n 3.0 onward, so Docker or Podman is required. Use npx n8n-node dev --external-n8n to attach to an n8n you already run.

Upgrading from 0.3.0 or earlier

Version 0.4.0 renames two identifiers that n8n stores in workflow JSON, before the node is verified for n8n Cloud and the names are fixed for good:

  • The chat memory sub-node's type is now redisAgentChatMemory (was redisAgentMemoryChatMemory). A workflow saved with an earlier version shows that node as unknown; delete it, add Redis Agent Chat Memory again and set its parameters.
  • The User ID parameter of Redis Agent Chat Memory and of Session → Add Event is stored as ownerId (was actorId), the key every other operation already uses. A saved value under the old key is ignored, so open those nodes and set User ID again.

The display names, the credential and the stored data are unchanged.

Create the credential

Both nodes use one credential, Redis Agent Memory API, with three fields:

Field Where to find it
Server URL AI Services / Iris / Agent Memory page, Configuration tab, General Settings. Enter the host only (https://…), without a path: the nodes add /iris/data/v1/agent-memory themselves.
Store ID Same page and tab.
API Key AI Services / Iris / Agent Memory page, API Keys tab.

The credential test calls the store's health endpoint, so a wrong URL, store or key fails at save time rather than at the first chat.

Summary views and extraction

Long-term memory extraction and summary views are features of the store. Extraction is on by default on hosted stores; summary views must be enabled on the store before the Summary View resource and the chat memory's Summary View ID option work.

Next: the Quick Start.