retainr

n8n community node for retainr.dev — AI agent memory persistence API. Store, search, and retrieve semantic memories for AI agents across workflow runs.

Package Information

Downloads: 0 weekly / 0 monthly
Latest Version: 0.2.7
Author: stackflo-labs

Available Nodes

Retainr icon
Retainr
Store, search, and retrieve AI agent memories.

Documentation

@stackflo-labs/n8n-nodes-retainr

n8n community node for retainr.dev — the AI agent memory persistence API.

Give your n8n AI agents long-term memory that persists across workflow runs. Store facts, preferences, and context, then retrieve them with semantic search — so your agents remember what matters.

Features

Operation Description
Store Save a memory with scope, tags, metadata, TTL, and deduplication
Search Semantic similarity search across stored memories
Get Context Retrieve pre-formatted memory context ready for LLM system prompts
List Browse memories with filters (scope, user, session, tags)
Delete Remove memories by filter criteria
Get Workspace Info View plan usage, API keys, and workspace details

Installation

n8n Cloud & Desktop

  1. Go to Settings > Community Nodes
  2. Search for @stackflo-labs/n8n-nodes-retainr
  3. Click Install

Self-hosted n8n

cd ~/.n8n
npm install @stackflo-labs/n8n-nodes-retainr

Then restart n8n.

Setup

  1. Sign up at retainr.dev and get your API key
  2. In n8n, go to Credentials > New > Retainr API
  3. Paste your API key (rec_live_...)
  4. The default Base URL (https://api.retainr.dev) works out of the box

Usage Examples

Store a memory after a conversation

  1. Add the Retainr node after your AI Agent
  2. Set Resource to Memory, Operation to Store
  3. Map the conversation summary to the Content field
  4. Set Scope to User and provide the User ID

Inject memory context into an LLM prompt

  1. Add a Retainr node before your AI Agent
  2. Set Operation to Get Context
  3. Use the incoming message as the Query
  4. Pass the context output into your LLM's system prompt

Deduplicate similar memories

Use the Dedup Threshold field (e.g., 0.95) when storing. If a sufficiently similar memory already exists, it will be updated instead of duplicated.

Memory Scopes

Scope Use case
session Single workflow run — requires session_id
user Persists across runs for one user — requires user_id
agent Shared across users for one agent — requires agent_id
global Shared across the entire workspace

AI Agent Tool

This node has usableAsTool enabled — you can use it directly as a tool inside n8n's AI Agent node, letting the agent decide when to store or recall memories autonomously.

API Documentation

Full API reference: retainr.dev/docs/api

License

MIT

Discussion