n8n: Prisma AIRS Runtime Security

The Prisma AIRS community node requires instance owner permissions for installation.

1. Open the n8n instance (Cloud or self-hosted).
2. Navigate to Settings > Community Nodes.
3. Confirm the Community Nodes page loads and the Install button is visible.

Runtime Security requires an active Prisma AIRS license, a configured AI Security Profile, and an API key.

1. Log in to Strata Cloud Manager.
2. Navigate to AI Access Security.
3. Confirm a Security Profile (AI Profile) exists and note its exact name (case-sensitive).
4. Obtain or generate an AIRS API key (this is the x-pan-token value).

The community node provides a native n8n interface for Prisma AIRS scan operations.

1. In n8n, navigate to Settings > Community Nodes.
2. Click Install a community node.
3. Enter @paloaltonetworks/n8n-nodes-prisma-airs in the search field (or visit the npm package page).
4. Click Install.
5. Restart the n8n instance if prompted.

The community node uses n8n's built-in credential system to store the API key and profile name securely.

1. In n8n, navigate to Credentials from the left-hand menu.
2. Click Add Credential.
3. Search for and select Prisma AIRS API.
4. Fill in the credential fields:

5. Click Save.

This step builds a minimal workflow that scans user prompts before forwarding them to an AI model. The community node handles API authentication and request formatting.

1. Create a new workflow in n8n.
2. Add a trigger node:
   • Chat Trigger for interactive chat workflows, or
   • Manual Trigger for testing.
3. Add a Prisma AIRS node after the trigger.
4. In the node properties:
   • Select the credential created in Step 2.1.
   • Set Operation to Prompt Scan.
   • Set Content to the expression referencing the trigger output (e.g., {{ $json.chatInput }}).
5. Add an IF node after the Prisma AIRS node.
6. Configure the IF condition:
   • Value 1: {{ $json.action }}
   • Operation: equals
   • Value 2: block
7. Connect the true (blocked) output to a Set node that returns a safe message (e.g., Content blocked by security policy).
8. Connect the false (allowed) output to the AI model node.

After the AI model generates a response, add a second scan to check the output for data leakage, malicious code, and policy violations before returning it to the user.

1. After the AI model node, add a Code node (optional) to clean the response text — normalize line endings, collapse excessive whitespace.
2. Add a second Prisma AIRS node.
3. In the node properties:
   • Select the same credential from Step 2.1.
   • Set Operation to Response Scan.
   • Set Content to the expression referencing the AI model output (e.g., {{ $json.output }}).
4. Add a second IF node to check the response scan verdict ({{ $json.action }} equals block).
5. Connect the true (blocked) output to a Set node returning Response blocked by security policy.
6. Connect the false (allowed) output to the final output node.

For deployments requiring custom headers, tr_id correlation, or non-US regional endpoints, use the HTTP Request node to call the Prisma AIRS Scan API directly.

Prompt Scan via HTTP Request

Add an HTTP Request node with the following configuration:

JSON body:

{
  "tr_id": "{{ $execution.id }}-prompt",
  "ai_profile": {
    "profile_name": "YOUR_PROFILE_NAME"
  },
  "metadata": {
    "app_user": "n8n-user",
    "ai_model": "{{ $json.ai_model }}",
    "application_name": "n8n-airs-integration"
  },
  "contents": [
    {
      "prompt": "{{ $json.prompt }}"
    }
  ]
}

Run the workflow with a benign prompt to confirm the end-to-end flow works.

1. Open the workflow in the n8n editor.
2. Click Execute Workflow (or send a chat message if using the Chat Trigger).
3. Provide a safe test prompt, e.g., What is the capital of France?
4. Observe the workflow execution: the Prisma AIRS prompt scan node should return action: allow.
5. The prompt flows to the AI model, generates a response, passes the response scan, and returns to the user.

Run the workflow with a prompt designed to trigger the security profile's block rules to confirm the block path works.

1. Send a prompt that matches one of the security profile's detection categories (e.g., a prompt containing sensitive data patterns or a known injection attempt).
2. Observe the workflow execution: the Prisma AIRS prompt scan node should return action: block.
3. The IF node routes to the block handler, which returns the safe message instead of forwarding to the AI model.

All scan events are recorded in Strata Cloud Manager for monitoring and compliance.

1. Log in to Strata Cloud Manager.
2. Navigate to AI Access Security > Activity Logs.
3. Locate the scan entries from the test workflow executions.
4. Confirm both the allowed and blocked scan results appear with the correct verdicts.