Fix Failing HTTP URLs with Gemini Diagnosis & Gmail Alerts

Fix failing HTTP URLs automatically—diagnose with Google Gemini and alert via Gmail

This n8n workflow tests a user-provided HTTP endpoint, captures the exact error details when it fails, uses a LangChain AI Agent powered by Google Gemini to diagnose and propose a corrected URL, and then sends a Gmail notification with the failure information.

What this workflow does

• Manual start: Run the workflow on demand using a manual trigger.
• Send an HTTP Request: The workflow sets the target URL (from the “HTTP URL” step) and issues an HTTP Request with the included latitude/longitude and weather “current” query parameters.
• Capture failure context: If the HTTP call fails, it extracts the URL, HTTP status, and the error message from the response.
• AI-powered diagnosis and fix: The error context is sent to a LangChain AI Agent running on the Google Gemini Chat Model, which returns structured JSON containing a diagnosis and a suggested corrected endpoint.
• Retry with the corrected URL: The workflow replaces the URL with the AI-suggested endpoint, waits, and then retries the request.
• Notify via Gmail: Finally, it sends a Gmail message containing the captured error code and error message for visibility and troubleshooting.

Use cases

• Debugging SaaS integrations when an HTTP endpoint starts returning errors.
• Monitoring critical API URLs and quickly getting an AI-suggested corrected endpoint.
• Assisting automation engineers during endpoint migrations or parameter changes.

Technical details

• n8n nodes: set, wait, gmail, sticky note, http request, manual trigger.
• AI: LangChain AI Agent using the Google Gemini Chat Model (requires Google Gemini (PaLM) API credentials).
• Email: Gmail with OAuth2 credentials (configure recipient, subject, and message format in the Gmail node).

Setup tip: Update the initial URL in the “HTTP URL” step to the endpoint you want to test and auto-correct.