Record Event
AI-generatedSummary
Records a standalone audit entry (actionKind "event") to Provedit's tamper-evident chain by POSTing to /v1/actions, once per input item. Use it to log things like "user said X" or "model returned Y" with no invocation/result follow-up.
Inputs
- Tool (required) — Tool name that your Provedit policy matches on (e.g. send_email, billing.refund).
- Params (JSON) — JSON object of parameters recorded on the action. Defaults to {}; invalid JSON throws an error.
- Target Kind — Category of the target: db, deploy, email, file, message, other, payment, record (default), repo, or url.
- Target Ref (required) — Identifier of what the action touched: a record ID, or a full http(s) URL for the URL kind.
- Target Summary — Human-readable line shown next to the target on the timeline.
- Additional Fields — Optional: Actor Agent Name, Actor Model Name, Actor User Email, Session ID (groups related actions), and Tool Version (default 0.0.0). Result Duration and Result Output apply only to Record Result.
Output shape
One output item per input item, containing the JSON response from the Provedit API for the created action. It is paired to the input item.
The exact response fields aren't defined in the node. The node sends no policy decision. The API re-evaluates the active policy and stamps its own decision. The actor vendor is always set to 'n8n'. With Continue On Fail enabled, failed items return {error: message} instead of stopping the run.
Examples
Example 1: Log that an AI agent drafted an email to a customer for the audit trail.
Tool=send_email, Target Kind=email, Target Ref=customer@example.com, Target Summary='draft reply', Additional: Actor Model Name=gpt-4o, Session ID={{$execution.id}}
Example 2: Record a refund-related event against a customer record.
Tool=billing.refund, Target Kind=payment, Target Ref=cust_123, Params={"amount":4800}, Target Summary='refund $48.00 to ACME Corp'