n8n Error Trigger to Telegram: Privacy-Safe Alerting

Get instant, privacy-safe Telegram alerts when your n8n workflows fail

This n8n automation workflow detects trigger and execution failures and sends actionable error notifications to Telegram—while automatically redacting sensitive data, classifying severity, and deduplicating repeated incidents.

What this workflow does

• Monitors errors automatically: It triggers whenever a linked n8n workflow fails during a trigger or execution.
• Normalizes and protects sensitive payloads: The error payload is normalized, then common secrets are redacted, including tokens, credentials, JWTs, and emails.
• Classifies incident severity: The workflow categorizes the error by severity/category so alerts can be filtered by policy.
• Prevents alert spam with deduplication: It generates a stable incident fingerprint and deduplicates repeated incidents within a configurable window.
• Applies alert policies: Alerts are sent only if they meet configured rules such as minimum severity, optional quiet hours, and workflow allow/block lists.
• Quietly suppresses when not allowed: If policy checks fail, the workflow suppresses the notification and outputs the suppression reason instead of alerting.
• Sends Telegram messages: When the policy allows it, it sends formatted incident details and includes an execution link when available.

Use cases

• Operations monitoring: Know immediately when an n8n workflow’s trigger fails or an execution crashes.
• SaaS reliability: Receive severity-based Telegram alerts while keeping secrets out of logs and messages.
• Team incident response: Reduce noise with quiet hours and deduplicated alerts for recurring errors.

Technical details

• Integrations/Nodes: error trigger, if, set, code, telegram, plus workflow notes via sticky note.
• Setup requirements: An n8n instance, a Telegram bot credential, and a destination chat/channel ID.
• Configuration includes: chat ID, optional environment/owner/timezone, quiet hours, minimum severity, and deduplication window.