n8n Error Workflow: Triage Failures with Bright Data, GPT & Slack

Instantly triage n8n failures with Bright Data, GPT, and Slack

If an n8n workflow crashes, this error workflow automatically gathers the failure details, checks the failing URL externally via Bright Data, consults the vendor’s Statuspage summary, and then uses OpenAI (GPT-5.6) to generate an evidence-based verdict—posted straight to the right Slack channel.

What this workflow does

• Triggers on errors: Uses the n8n Error Trigger when any other workflow execution fails.
• Extracts failure evidence: Pulls the failing node, error message, HTTP code, and the URL/host involved, then classifies the failure (e.g., auth, rate limit, connectivity, server error, or internal code).
• Validates external availability with Bright Data: When the failure involves an external (non-owned) host, it calls Bright Data Web Unlocker to fetch the URL from an external IP and derives the vendor’s Statuspage summary URL.
• Detects active incidents: Fetches /api/v2/summary.json through Bright Data and checks whether the vendor has declared an active incident.
• Decides what likely broke: Combines the available evidence to distinguish between a vendor outage, a rate limit, or a potential bug on your side.
• Generates a short, actionable verdict: Uses OpenAI (GPT-5.6) to produce a two-sentence explanation of what broke and what to do next.
• Routes to the right team in Slack: Posts the verdict, evidence, and execution link to different Slack channels (incidents, ops, or dev) based on the verdict.

Use cases

• Speeding up incident response when an external API change or outage causes n8n workflows to fail.
• Separating “our bug” vs “vendor is down” by verifying vendor status via Bright Data.
• Reducing noise in ops/developer channels by automatically flagging rate limits and connectivity issues.

Technical details

• Integrations: Bright Data (Web Unlocker via https://api.brightdata.com/request) and Slack.
• AI: OpenAI (GPT-5.6) for a two-sentence evidence-based verdict.
• n8n nodes mentioned: if, set, code, no op, merge, slack.
• Setup: Add a Bright Data HTTP Header Auth credential for the Bright Data request endpoint and configure the Bright Data zone name; add an OpenAI credential and reference the specified chat model.