Docs/n8n Workflows

SarvCrawl / Docs /n8n Workflows

SarvCrawl for n8n

Drop the SarvCrawl node into any n8n workflow to scrape, crawl, search, and manage knowledge bases — no code, just wires.

n8n-nodes-sarvcrawl

New to n8n? n8n is a visual workflow automation tool. You connect nodes on a canvas — each node does one thing, and data flows along the wires from left to right.

What is the SarvCrawl node?

A single community node that wraps the entire SarvCrawl API. Add it to a workflow, pick a Resource and an Operation, and it calls your backend for you.

A SarvCrawl node sits in the middle of any workflow

Schedule Trigger

main

SarvCrawl

main

Send to Slack

Manage KBs

Create, list, fetch, and delete knowledge bases.

Ingest content

Scrape a URL or crawl an entire website into a KB.

Full-text search across everything you have stored.

Installing the node

The fastest way for any user — install straight from n8n's Community Nodes panel. No cloning, no files.

Open Community Nodes

In n8n, go to Settings → Community Nodes and click Install.

Enter the package name

Type n8n-nodes-sarvcrawl, accept the risk prompt, and confirm.

Find it in the node panel

Press Tab on the canvas and search for “SarvCrawl”. Drag it in.

Self-hosted n8n must have community packages enabled (the default). Prefer the CLI? See for the developer install.

Set up credentials

The node authenticates to your SarvCrawl backend with an API key. Create the credential once and reuse it on every SarvCrawl node.

SarvCrawl

Parameters

SarvCrawl API credential

API Key

kb-••••••••••••••••••••

Sent as the x-api-key header on every request.

Base URL

https://crawl.sarv.com

Where your SarvCrawl backend runs.

Connection tested against GET /auth/me

How to add it

On any SarvCrawl node, open the Credential dropdown → Create New.
Paste your API Key from SarvCrawl settings.
Set the Base URL to your backend address.
Click Test — a green check means you’re connected.

Anatomy of a node

Every n8n node has an input handle on the left, an output handle on the right, and a parameter panel you open by double-clicking it.

SarvCrawl

Input handle — wire data in from the previous node.

Output handle — the node’s result flows out to the next.

The two dropdowns that drive everything

Open the panel and you’ll always start with the same two fields. Resource picks what you’re working with; Operation picks the action. The rest of the fields change to match.

SarvCrawl

Parameters

Resource

Job

Operation

Scrape URL

Knowledge Base ID

kb_abc123

URL

https://example.com/article

Include Linked Documents

true

Resources & Operations

Three resources, ten operations. Here is everything the node can do and the fields each operation needs.

Knowledge Base

Create and manage the containers that hold your scraped content.

Operation	What it does	Key fields
`Create`	Create a new knowledge base	Name, Description
`Get`	Fetch a single KB by ID	Knowledge Base ID
`Get All`	List every KB you own	—
`Delete`	Permanently delete a KB and its contents	Knowledge Base ID

Job

Ingest content and track the async jobs that process it.

Operation	What it does	Key fields
`Scrape URL`	Scrape one URL into a KB	KB ID, URL, Include Docs
`Crawl Website`	Crawl a whole site from a root URL	KB ID, URL, Max Depth, Page Limit
`Get Status`	Poll a job until it is completed	Job ID
`Get All Jobs`	List all jobs for a KB	Knowledge Base ID
`Cancel`	Stop a running or queued job	Job ID

Keyword or semantic search across everything stored in a KB.

Operation	What it does	Key fields
`Search`	Paginated full-text (keyword/BM25) search over a KB	KB ID, Query, Page, Size, Filter
`Semantic Search`	Hybrid embedding + BM25 (RRF) chunk-level search over a KB	KB ID, Query, Size, Filter

Connecting nodes

A workflow is just nodes wired together. Drag from one node's output handle to the next node's input handle — data flows along the wire.

Start with a trigger

Every workflow begins with a trigger node (Manual, Schedule, Webhook…). It has no input handle — it’s where execution starts.

Drag a wire to SarvCrawl

Click the trigger’s right-hand output handle and drag to the SarvCrawl node’s left input handle.

Chain more nodes

The SarvCrawl node’s output (e.g. a job_id) becomes the input of whatever comes next — another SarvCrawl node, an IF, a Slack message.

Each wire carries the previous node's output forward

Manual Trigger

main

Create KB

main

Scrape URL

main

Get Status

Passing data between nodes

Use n8n expressions to feed one node's output into the next. Wrap a reference in {{ }} and it resolves at run time.

After a Create KB node returns { id }, reference it in the next node with an expression instead of hard-coding the ID:

SarvCrawl

Parameters

Resource

Job

Operation

Scrape URL

Knowledge Base ID

Pulls the id from the previous node's output.

URL

https://example.com

Handy references

{{ $json.id }}

A field from the previous node

{{ $json.job_id }}

The job id returned by Scrape/Crawl

{{ $node['Create KB'].json.id }}

A field from a specific named node

{{ $json.hits[0].url }}

First search hit’s URL

Example · Scrape, poll, then search

The bread-and-butter pattern: ingest a page, wait until the async job finishes, then query what was stored.

Manual Trigger

main

Scrape URL

main

Wait 30s

main

Get Status

main

Scrape returns a job_id immediately. The Wait node gives the worker time to finish, then Get Status confirms completed before you Search the stored content.

{
  "resource": "job",
  "operation": "getStatus",
  "jobId": "={{ $json.job_id }}"
}

Example · Nightly site crawl

Keep a knowledge base fresh: crawl a docs site on a schedule and post a summary to Slack when it's done.

Schedule (2am)

main

Crawl Website

main

Wait 5m

main

Get Status

main

Slack message

The Schedule Trigger fires nightly. Crawl Website kicks off a full crawl; after a wait, Get Status reports how many pages were stored, which the Slack node drops into your channel.

{
  "resource": "job",
  "operation": "crawl",
  "kbId": "kb_docs",
  "url": "https://docs.example.com",
  "maxDepth": 5,
  "limit": 500,
  "includeDocs": true
}

Example · Webhook RAG endpoint

Turn SarvCrawl into a question-answering API: a webhook takes a question, searches the KB, an AI node answers from the hits, and the webhook responds.

Webhook

main

AI Answer

main

Respond

The Webhook receives { question }. The Search node queries the KB with {{ $json.body.question }}, the AI node grounds its answer on the returned hits, and Respond to Webhook returns it to the caller.

{
  "resource": "search",
  "operation": "search",
  "kbId": "kb_support",
  "query": "={{ $json.body.question }}",
  "size": 5,
  "filterType": ""
}

Local testing (for developers)

Building the node from source? Link it into n8n's custom-extensions folder and iterate with the TypeScript watcher.

# From the n8n-nodes-sarvcrawl folder
npm install
npm run build      # compiles TypeScript → dist/
npm link

# Link into n8n's custom-extensions folder
mkdir -p ~/.n8n/custom
cd ~/.n8n/custom
npm init -y        # first time only
npm link n8n-nodes-sarvcrawl

# Start n8n — it auto-loads nodes from ~/.n8n/custom
n8n start

Open http://localhost:5678, add the SarvCrawl node, create a SarvCrawl API credential pointing at your backend, and run the quick-test workflow from the package README.

MCP Server All Docs