Low-confidence AI answer monitor webhook template

How it works

1. 1.A webhook node accepts POST requests at path ai-agents-low-confidence-answer-monitor-form with response deferred to a response node.
2. 2.A Set node adds source (fixed string) and input (request body) fields while keeping original fields.
3. 3.An If node checks that input is not empty and routes the flow accordingly.
4. 4.With input present, Create CRM Record and then Notify Owner call placeholder HTTP endpoints using genericCredentialType authentication.
5. 5.Respond to Webhook nodes finish with a JSON status of created or missing-fields.

Use cases

• Logging low-confidence AI agent answers into a CRM for human review.
• Notifying owners to review uncertain answers before they are used.

Requirements

• Enable the webhook and send POST requests to the configured path.
• Configure genericCredentialType credentials for both HTTP nodes and replace the api.example.com URLs and placeholder values with real endpoints and payloads.

Setup and testing in n8n

1. POST a body with data to the webhook path and confirm both placeholder HTTP calls run and the response is created.
2. POST an empty body and confirm the response is missing-fields with no HTTP nodes executed.

Service Value

Ready-made workflow template for automation delivery and service execution.