Validate
AI-generatedSummary
Validate a user prompt against Akto guardrails and separate allowed from blocked results.
Inputs
- prompt (required) — The prompt to validate, sourced from body.chatInput or chatInput.
Output shape
An array of two result arrays: the first contains items marked as allowed, the second contains items marked as blocked; each item includes allowed, blocked, reason, webhookHost, webhookPath, clientIp, chatInput.
When using Validate operation the node returns a two‑element array; the first element is the list of allowed items, the second the list of blocked items. Individual items contain boolean flags and metadata from the guardrails response.
Examples
Example 1: Validate a prompt using Akto guardrails
Operation = Validate, provide Prompt, use AktoApi credential