engram0

n8n node to add and search memories for AI agents via Engram0 API

Package Information

Downloads: 24 weekly / 624 monthly
Latest Version: 0.3.0
Author: Engram0

Available Nodes

Engram0 icon
Engram0
Add or search memories for AI agents via Engram0 API

Documentation

n8n-nodes-engram0

n8n community node to add and search memories for AI agents using the Engram0 API.

Use it in n8n workflows to give your agents long-term memory: store what users say or what the agent learns, then search relevant memories before replying.

Requirements

  • n8n (self-hosted or n8n Cloud)
  • Engram0 API key (create one in your Engram0 dashboard under API Keys, or use a self-hosted compatible API)

Installation

  1. In n8n: Settings → Community nodes → Install a community node
  2. Enter: n8n-nodes-engram0
  3. Install and restart if prompted

Or install via CLI in your n8n user directory:

npm install n8n-nodes-engram0

Credentials

Create an Engram0 API credential:

  • API Key: Your Engram0 API key (e.g. l3_xxxx... from the dashboard).
  • Base URL: https://api.engram0.com for Engram0, or your self-hosted API URL.

Node: Engram0

Operations

  • Add Memory – Store a message or fact as memory. Use after a user message or when the agent decides to remember something.
  • Search Memories – Search by query and get relevant memories. Use before generating a response to inject context.

Scoping (User ID / Agent ID / Run ID)

At least one of these is required for Add Memory so memories are scoped correctly:

  • User ID: End-user identity (e.g. your app’s user id).
  • Agent ID: Which bot or assistant (e.g. support-bot).
  • Run ID: A specific conversation or session.

Use the same IDs in both Add and Search so context stays consistent.

Example workflow

  1. Trigger (e.g. webhook or chat trigger) receives a user message.
  2. Engram0 – Add Memory: Store the user message with User ID (and optionally Agent ID / Run ID).
  3. Engram0 – Search Memories: Query with the user message (or a summarized version) and same User ID.
  4. Use the Search results in your LLM node (e.g. inject into system prompt or context), then generate the reply.

API compatibility

Works with:

  • Engram0
  • Any self-hosted API compatible with the same endpoints

Uses POST /memories and POST /search with header X-API-Key. See Engram0 for API details.

License

MIT

Discussion