Privent is a native n8n node. Drag it into any workflow canvas between your data nodes and LLM nodes. It inspects node input/output JSON, applies policy, and passes a safe payload downstream. No proxy. No code. Policy defined per workflow.
Workflow data flows to LLM
Your n8n workflow pulls data from databases, APIs, or HTTP nodes and feeds it into an LLM node. The data is assembled automatically; no human review happens before it reaches the external model.
Privent node intercepts
The Privent node sits between your data source and the LLM node. It reads the full node input JSON, structured or unstructured, and runs ACARS scoring against your deployment group's policy.
Safe payload forwarded
APE transforms sensitive fields inline. The downstream LLM node receives a clean payload. Workflow execution continues. Every transformation is logged to the Privent audit trail.
Add Privent to your existing n8n pipeline without changing your architecture. One addition. Full graph-state visibility.
{"nodes": [{ "name": "HTTP Request", "type": "n8n-nodes-base.httpRequest" },{"name": "Privent", // ← drag & drop"type": "n8n-nodes-privent.privent","parameters": { "policy": "default", "deploymentGroup": "ops" }},{ "name": "OpenAI", "type": "n8n-nodes-base.openAi" }]}
External gateways receive only the final prompt string. Privent reads everything in context, in real time.
Intercepts structured database output (customer records, financial data) before it's fed into an OpenAI prompt node.
Detects API keys, tokens, or auth headers that accidentally appear in workflow data passed to external LLM nodes.
Apply different risk policies per workflow: finance workflows get stricter thresholds than marketing automation workflows.
We integrate in under 30 minutes. No orchestration changes required. Your pipelines keep running. Privent keeps watching.
