n8n Workflow: Toxicity Scoring from Form Submissions (GPT-4o Mini)

Automatically score message toxicity from n8n Form submissions—with a GPT-4o Mini HTML report on the completion page

This n8n workflow analyzes every client message submitted through an n8n Form, generates a 0–100 toxicity score using OpenAI (GPT-4o Mini), and displays a styled HTML report (with verdict, flags, and protection tips) directly on the form completion page.

What this workflow does

• Receives a client message or project brief via an n8n Form Trigger.
• Sends the text to OpenAI (GPT-4o Mini) to return strict JSON containing:
  • a 0–100 toxicity score
  • a verdict
  • red/yellow/green flags
  • protection tips aligned to the verdict
• Parses the AI’s JSON and renders a color-coded HTML report (score, overall verdict, flags, and tips).
• Displays the report immediately on the n8n Form completion page.

Use cases

• Screening inbound contact form messages for toxic language before it reaches your team.
• Adding a safety layer to lead intake for SaaS operators and automation engineers.
• Reviewing red/yellow/green warning signs for higher-quality support and onboarding conversations.
• Generating a consistent, shareable HTML summary for moderation or internal follow-up.

Technical details

• Form Trigger to capture client input from an n8n Form
• OpenAI (GPT-4o Mini) via an LLM node (n8nn8n-nodes-langchainchain / n8nn8n-nodes-langchainlm chat open ai) to return strict JSON
• Code step to parse JSON and generate the styled HTML report
• Optional customization of the analysis prompt, verdict thresholds, and HTML styling

Get started by adding your OpenAI API credential, activating the workflow, and opening the Form Trigger’s Production URL to see the report on completion.