n8n-nodes-rag-p

This is an n8n community node for RAG-Platform (rag-p) — a managed Retrieval-Augmented Generation service.

It lets you query your datasets, upload documents, and check dataset status from any n8n workflow.

Installation · Operations · Credentials · Example workflow · Resources

Installation

Follow the installation guide in the n8n community nodes documentation.

In short:

1. Open your n8n instance.
2. Go to Settings → Community Nodes.
3. Select Install and enter n8n-nodes-rag-p.
4. Agree to the risks of using community nodes and confirm.

After installation the RAG-Platform node will appear in the nodes panel.

Manual install (self-hosted)

cd ~/.n8n/custom
npm install n8n-nodes-rag-p

Restart n8n.

Credentials

Create a credential of type RAG-Platform API.

Field	Description
API Key	Personal API key from https://lekottt.ru/dashboard/api-keys
Base URL	API base URL. Default https://api.lekottt.ru. Override for self-host
Verify SSL	Toggle off only for development with self-signed certificates

The credential is verified automatically against GET /api/v1/auth/me.

Operations

The node currently supports three operations on the Dataset resource.

Query

POST /api/v1/rag/query — run a RAG query against a dataset.

Parameter	Type	Description
Dataset ID	string	Target dataset id
Query	string	Natural-language question
Top K	number	Number of chunks to retrieve (1..50)

Returns: { answer, citations, usage }.

Upload Document

POST /api/v1/datasets/{dataset_id}/documents — upload a document into a dataset.

Parameter	Type	Description
Dataset ID	string	Target dataset id
Input Type	text/binary	Whether to upload raw text or a binary file
Text Content	string	(when text) Plain-text body
Binary Property Name	string	(when binary) Property name on the input item
Filename	string, optional	Display name in the dataset

Returns: { document_id, chunks_count, status }.

Get Dataset

GET /api/v1/datasets/{dataset_id} — read dataset metadata, useful for polling indexing status.

Returns: { id, name, documents_count, indexed_status }.

Example workflow

A simple Q&A bot:

1. Trigger: Telegram message received.
2. RAG-Platform → operation Query, with query mapped from the Telegram message text.
3. Telegram → send {{ $json.answer }} back to the chat.

For ingestion:

1. Trigger: Google Drive new file in folder.
2. RAG-Platform → operation Upload Document, with binary input from the previous node.
3. Wait + RAG-Platform → operation Get Dataset, until indexed_status == "ready".

Resources

• n8n community nodes documentation
• rag-p API reference
• Issues

Publishing

Releases are automated via the n8n-publish.yml
GitHub Actions workflow. It runs npm ci && npm run build && npm publish --access public
from integrations/n8n/ against the public npm registry.

One-time setup

Add an npm automation token (with publish rights on n8n-nodes-rag-p) as the
repository secret NPM_TOKEN:

gh secret set NPM_TOKEN --body "<npm-automation-token>"

Releasing a new version

1. Bump version in package.json and merge the change to
   main.

2. Trigger the workflow either manually or via a tag:

   Manual run:

   gh workflow run n8n-publish.yml
   

   Tag-driven release (pushes matching n8n-v* trigger automatically):

   git tag n8n-v0.1.1
   git push origin n8n-v0.1.1
   

The workflow itself does not bump the version — it publishes whatever is in
integrations/n8n/package.json on the checked-out ref.

License

MIT