ChatGPT AI Agent for Search Console Data (n8n + Postgres)

Chat with your Google Search Console data using an n8n AI Agent—powered by OpenAI + Postgres

This n8n workflow turns your Google Search Console reports into a conversational experience. Instead of manually digging through Search Console screens, you can ask questions in chat and have the agent respond using your Search Console data—while storing/organizing information in Postgres.

What this workflow does

• Provides an AI Agent inside n8n that you can use as a chat interface to query your Search Console data.
• Includes node-by-node documentation within the template so you can configure setup and understand how each step works.
• Requires you to configure Google Search Console OAuth credentials so the workflow can make Search Console API calls.
• Uses a tested/optimized system prompt for compatibility—note that an earlier original prompt was not effective with gpt4-o-mini, and the workflow includes prompts you can test from the provided Notion page (including one that works well with gpt4-o-mini).
• Highlights an important setup detail: you must set the correct OAuth scopes to avoid issues with refresh tokens and frequent re-authentication.

Use cases

• Quickly answer questions like “Which pages saw the biggest clicks increase last week?” without leaving your workflow.
• Search Console reporting support for SaaS operators who want faster insights for clients.
• Automation engineers building an internal “SEO data assistant” using n8n and OpenAI with persistent storage in Postgres.

Technical details

• AI + OpenAI (including updated prompt compatibility for gpt4-o-mini)
• Google Search Console via OAuth credentials (configure per n8n OAuth guidance)
• Postgres for storing/working with data
• n8n nodes used: set, switch, webhook, aggregate, sticky note, http request