easybits

n8n node for running easybits pipelines in workflows

Package Information

Released: this month
Downloads: 0 weekly / 0 monthly
Latest Version: 0.1.2
Author: Ivan Afonchenko

Documentation

@easybits/n8n-nodes-easybits

npm version
License: MIT

n8n community node that runs an easybits pipeline from your workflows.

An easybits pipeline is a sequence of processing steps (AI text generation with Gemini, OpenAI, Anthropic or Nebius, OCR, image generation, PII redaction, templating…) that you build in the easybits admin panel. This node sends text and optional files to the pipeline's API channel and returns the result.

Installation

Follow the n8n community nodes installation guide.

The package name is: @easybits/n8n-nodes-easybits

Configuration

Credential

Each credential points at one pipeline, through one of its API channels. Find the channel ID and API key on the pipeline's page in the easybits admin panel.

Credential Field Description
Channel ID The ID of the pipeline's API channel
API Key The channel's API key (stored as a secret)
Base URL The easybits API Controller. Default: https://ctrl-eb-lite.easybits.tech. Change it only for a self-hosted or development deployment.

The credential's connection test checks the channel ID and API key without running the pipeline.

Node parameters

Parameter Description
Content The text sent to the pipeline. Required. For an AI module it is the user message.
Files None (default), Binary Files, Data URLs, or Auto (Both)
Data URL Field JSON field with base64 Data URL(s). Shown for Data URLs and Auto. Supports dot notation (response.attachments.0.url). Default: dataUrl

Options

Option Default Description
Simplify Output on Return only the pipeline's final output. Off returns the full response, including the step trace.
Timeout (Seconds) 300 How long to wait for the pipeline to finish

Which file types a pipeline accepts depends on its modules. A request — content plus files — is limited to 5 MB.

How it works

The node runs the pipeline once per input item:

  1. Reads the item's content and, depending on Files, its binary attachments and/or Data URLs (binary files are converted to Data URLs)
  2. POSTs { "content": "...", "files": [...] } to {Base URL}/channel/{Channel ID} with the API key as a Bearer token
  3. Waits for the pipeline to finish and returns its output as one item

Output

With Simplify Output on:

  • If the pipeline's final output is a JSON object (e.g. an AI step with a structured output schema), its fields become the item's fields:
    { "invoice_number": "INV-001", "total": 120.5 }
    
  • Otherwise the output is returned as text:
    { "text": "The summary of the document is ..." }
    
  • Output files that are Data URLs (e.g. generated images) become binary data on the item: file_0, file_1, …

With Simplify Output off, the item is the API response as-is: { "messages": [...] }. The final output is messages[messages.length - 1].output.

Errors

Pipeline errors are shown with their easybits error code and the failing module and step, for example MAX_TOKENS (the model hit its output token limit) or STEP_TIMEOUT. With Continue On Fail enabled, a failed item outputs { "error": "..." } and the remaining items still run.

Usage

Run text through a pipeline

[Trigger] → [easybits]
  • Content: {{ $json.message }}
  • Files: None

Send a document

[Read Binary File] → [easybits]
  • Content: an instruction, e.g. Extract the invoice data
  • Files: Binary Files

Send a Data URL from a previous step

[HTTP Request / Set] → [easybits]
  • Files: Data URLs
  • Data URL Field: dataUrl
{ "dataUrl": "data:image/png;base64,iVBORw0KGgo..." }

The field can also hold an array of Data URLs.

Use it as an AI agent tool

The node can be attached to an n8n AI Agent as a tool, letting the agent call your pipeline.

Compatibility

Requires n8n v1.0 or later.

Resources

License

MIT

Discussion