Description

n8n-nodes-goodmem

This is an n8n community node that integrates with Goodmem, a semantic memory and knowledge storage platform. It enables n8n workflows to store, retrieve, and manage semantic memories with powerful embedding and search capabilities.

!n8n.io – Workflow Automation

Features

• Spaces Management – Create and delete memory spaces (containers for organizing memories)
• Memory Operations – Full CRUD operations for storing and retrieving memories
• Semantic Search – Retrieve memories using natural language queries with relevance scoring
• Flexible Chunking – Configure recursive, sentence-based, or custom chunking strategies
• Advanced Retrieval – Post-processing with reranking, LLM generation, and relevance filtering
• Metadata Support – Tag memories and spaces with custom labels and metadata for filtering

Table of Contents

• Prerequisites
• Installation
• Credentials
• Operations
• Usage Examples
• Chunking Strategies
• Advanced Retrieval Options
• Development
• Resources
• License

Prerequisites

• n8n version 1.60.0 or later
• A Goodmem account with API access
• Goodmem API key and server URL

Installation

Goodmem Installation

Configure a goodmem server to connect to before using the node.
Visit https://goodmem.ai/quick-start for more instructions. Consider using the Railway or Fly.io installation for quick access to the Goodmem API over the internet.

Community Nodes (Recommended)

1. Open your n8n instance
2. Go to Settings > Community Nodes
3. Select Install
4. Enter @pairsystems/n8n-nodes-goodmem and click Install

Manual Installation

cd ~/.n8n/nodes
npm install @pairsystems/n8n-nodes-goodmem

Then restart your n8n instance.

Credentials

Before using the node, configure your Goodmem credentials:

1. In n8n, go to Credentials > New
2. Search for Goodmem API
3. Enter the following:
– Goodmem Server: Your Goodmem API server URL
– Goodmem API Key: Your API key for authentication

Operations

Spaces

| Operation | Description |
|———–|————-|
| Create | Create a new semantic memory space with custom embedders and chunking configuration |
| Delete | Remove an existing space and all its memories |

Memories

| Operation | Description |
|———–|————-|
| Create | Store new content as a memory with optional chunking and metadata |
| Get | Retrieve a memory’s metadata and optionally its content and processing history |
| Delete | Remove a specific memory from a space |
| Download Content | Retrieve the original content of a stored memory |
| Retrieve | Perform semantic search across one or more spaces with natural language |

Usage Examples

Creating a Memory Space

Configure a space to store knowledge base documents:

1. Add the Goodmem node to your workflow
2. Select Space as the resource and Create as the operation
3. Enter a Space Name (e.g., “Product Documentation”)
4. Configure an embedder with ID and retrieval weight
5. Optionally set labels for organization (e.g., environment: production)

Storing a Memory

Add content to your space:

1. Select Memory as the resource and Create as the operation
2. Provide the Space ID where the memory will be stored
3. Enter the Content (text, markdown, etc.)
4. Optionally configure chunking strategy:
– Recursive: Splits by separators with configurable chunk size and overlap
– Sentence: Splits on sentence boundaries
– None: Store content as a single chunk

Semantic Search

Retrieve relevant memories using natural language:

1. Select Memory as the resource and Retrieve as the operation
2. Enter your Message (search query)
3. Optionally filter by:
– Specific Space IDs
– Filter expressions for metadata matching
– Relevance threshold (0-1)
4. Configure post-processing:
– Reranker: Improve result relevance ordering
– LLM Generation: Generate responses based on retrieved context

Chunking Strategies

The node supports flexible chunking for optimal embedding and retrieval:

Recursive Chunking (Default)

Best for general-purpose document processing:

• Chunk Size: Target size for each chunk (default: 512)
• Chunk Overlap: Overlap between chunks for context continuity (default: 64)
• Separators: Custom separators for splitting (paragraphs, sentences, etc.)
• Length Unit: Characters or tokens

Sentence-Based Chunking

Preserves sentence boundaries:

• Max Chunk Size: Maximum characters/tokens per chunk
• Min Chunk Size: Minimum chunk size threshold

Custom JSON

For advanced configurations, provide a custom chunking strategy as JSON.

Advanced Retrieval Options

Fine-tune search results with post-processing:

| Option | Description |
|——–|————-|
| Reranker ID | UUID of a reranker model to improve result ordering |
| LLM ID | UUID of an LLM to generate contextual responses |
| Relevance Threshold | Minimum score (0-1) for including results |
| LLM Temperature | Creativity setting for LLM generation (0-2) |
| Max Results | Limit the number of returned memories |
| Chronological Resort | Reorder results by creation time |

Example Workflow: AI-Powered Q&A System

[Webhook Trigger] --> [Goodmem: Retrieve] --> [OpenAI: Generate Response] --> [Respond to Webhook]

1. Webhook Trigger: Receives a user question
2. Goodmem Retrieve: Searches semantic memories for relevant context
3. OpenAI: Generates an answer using the retrieved context
4. Response: Returns the AI-generated answer

Development

Building

npm install
npm run build

Linting

npm run lint
npm run lint:fix

Local Testing

Link the node to your local n8n installation:

npm link
cd ~/.n8n/nodes
npm link @pairsystems/n8n-nodes-goodmem

Compatibility

Compatible with n8n version 1.60.0 or later.

Resources

• n8n Community Nodes Documentation
• Goodmem Documentation

License

MIT

Author

Zaid Abdulrehman (zaid.abdulrehman@pairsys.ai)