Skip to main content
Platform
Integrations
n8n
Example Workflows

Example n8n Workflows

These patterns show how the NordStellar node fits together with the rest of n8n. Each one lists the nodes in order along with the values and expressions that matter — adapt the destinations and thresholds to your own environment.

All examples assume you have completed Setup and have a project ID to hand. For the full parameter list of each operation, see the Node Reference.

Example 1: Post New Findings to Slack on a Schedule

The most common pattern: poll for events added since the last run and send them to a chat channel. This gives you SIEM-style alert routing without a SIEM.

Schedule Trigger

Set the interval to run every hour. The interval you choose here determines the lookback window in the next node.

NordStellar — Event: Get Many

  • Resource: Event

  • Operation: Get Many

  • Project ID: your project GUID

  • Return All: on

  • Filters → Date Added From:

    {{ $now.minus({ hours: 1 }).toISO() }}

Leave Date Added To empty so the query runs up to the present moment. Because the lookback matches the schedule interval, each event is picked up exactly once.

If — filter by severity

Branch on the event's risk level so routine findings don't page anyone:

{{ $json.risk_level }}

Route only HIGH and CRITICAL values down the alerting branch. You can send the remainder to a digest or drop them.

Slack — Send Message

Compose the message from the event fields:

New NordStellar finding: {{ $json.type }}
Risk: {{ $json.risk_level }}
Detected: {{ $json.date_added }}
Event ID: {{ $json.id }}
ℹ️

A fixed lookback window is simple and works well, but a failed or delayed execution can skip events. If you need stronger guarantees, store the timestamp of the last successful run — in n8n static data, a database, or a spreadsheet — and use that value in Date Added From instead.

Example 2: Enrich a Finding With Its Full Details

Get Many returns summary records. When you need the substance of a finding, such as the credentials captured by an info-stealer, chain a Get onto it.

NordStellar — Event: Get Many

Configure as in Example 1 to produce a stream of event summaries.

Loop Over Items

Add a Loop Over Items node so each summary is processed individually and one failure doesn't stop the batch.

NordStellar — Event: Get

  • Resource: Event

  • Operation: Get

  • Event Type: set from the summary using an expression:

    {{ $json.type }}
  • Event ID:

    {{ $json.id }}

The type value on a summary record already matches the API values the Get operation expects, so it can be passed straight through.

If your project produces event types you don't want to expand, put a Filter node before the Get step. Fetching details for every event in a large result set multiplies your API calls by the number of events.

Example 3: Create a Jira Ticket and Resolve the Event on Close

A two-way pattern that keeps NordStellar and your ticketing system in step.

Part one — open a ticket for each finding

  • Schedule Trigger — run on your preferred interval.

  • NordStellar — Event: Get Many — pull recent events, as in Example 1.

  • Jira — Create Issue — build the summary and description from the event fields. Include the NordStellar event ID in the description or a custom field so you can find it again:

    NordStellar event ID: {{ $json.id }}

Part two — resolve the event when the ticket closes

  • Jira Trigger — listen for the issue closed event.
  • Code or Set — extract the NordStellar event ID from the issue description or custom field.
  • NordStellar — Event: Resolve
    • Event ID: the extracted GUID
    • Resolved: on

Reopening works the same way: trigger on the issue being reopened and run Resolve with Resolved turned off.

Example 4: Forward Findings to a SIEM Without a Native Connector

NordStellar ships native connectors for Microsoft Sentinel and other SIEM platforms. If your SIEM isn't among them, n8n can bridge the gap.

Schedule Trigger

Run every 15 minutes, or whatever ingestion latency your SOC accepts.

NordStellar — Event: Get Many

  • Return All: on

  • Filters → Date Added From:

    {{ $now.minus({ minutes: 15 }).toISO() }}

Set — shape the payload

Map the NordStellar fields onto whatever schema your SIEM expects, adding any static fields such as a source name or index.

HTTP Request — send to your SIEM

Point the request at your SIEM's ingestion endpoint — for example a Splunk HTTP Event Collector or an Elastic data stream — using the authentication that platform requires.

ℹ️

Use a separate n8n credential for the SIEM endpoint. Your NordStellar credential cannot be used for it: the credential refuses to attach your access token to any host outside nordstellar.com.

Example 5: Give an AI Agent Access to NordStellar

Because the node is usable as a tool, an n8n AI Agent can query NordStellar while working through an investigation.

Add an AI Agent node

Configure it with the chat model of your choice.

Attach NordStellar nodes as tools

Connect one or more NordStellar nodes to the agent's Tool input. A useful starting set is:

  • Project → Get Many — so the agent can discover available projects.
  • Event → Get Many — so it can list recent findings.
  • Event → Get — so it can pull detail on anything interesting.

Leave the parameters the agent should control — such as Project ID or Event ID — set to be filled by the model.

Consider withholding Resolve

Attaching Event → Resolve lets the agent change state in NordStellar. Leave it out unless you specifically want that, and put a human approval step in front of it if you do.

⚠️

On self-hosted n8n, using community nodes as agent tools requires the environment variable N8N_COMMUNITY_PACKAGES_ALLOW_TOOL_USAGE=true. Review what an agent is permitted to do before enabling it against production data — actions it takes through the node are performed with your access token and are attributed to you.

Tips for Reliable Workflows

  • Match your lookback to your schedule. If the workflow runs hourly, query the last hour. Overlapping windows create duplicate alerts; gaps lose findings.
  • Use Return All deliberately. It pages through the entire result set, which is what you want for a 15-minute window and rarely what you want for a six-month one.
  • Handle partial failures. Enable Continue On Fail on nodes that process batches so one bad item doesn't abort the execution.
  • Keep tokens in credentials. Reference the n8n credential rather than pasting an access token into an HTTP Request node or a Code node.
  • Test with a narrow scope first. Point a new workflow at a single project and a short date range before letting it run across your whole organization.

If you run into problems, the Troubleshooting section of the setup guide covers the most common causes.

NordStellar © 2026Privacy Policy