n8n-monitoring（n8n 监控）

n8n Monitor Skill

Monitor your self-hosted n8n instance through a webhook-based API. Get health reports, detect failing workflows, and debug errors using structured data from your n8n executions.

Quick Workflow

1. Call the webhook with {"action": "get_workflow_executions", "limit": 50} to get a health overview
2. Check the summary.failureRate to determine status: GREEN (<10%), YELLOW (10-25%), RED (>25%)
3. If failures exist, check workflowPerformance.topProblematicWorkflows for the worst offenders
4. For deep debugging, call {"action": "get_execution_details", "limit": 5, "workflow_id": "<ID>"} to get error messages and failed nodes

Available Endpoints

All endpoints are POST requests to the N8N_WEBHOOK_URL environment variable.

get_active_workflows

Returns all active workflows with IDs, names, and metadata.

{"action": "get_active_workflows"}

get_workflow_executions

Returns recent executions with computed KPIs: failure rate, timing metrics, per-workflow performance, alerts.

{"action": "get_workflow_executions", "limit": 50}

Key response fields:

• summary.totalExecutions, summary.failureRate, summary.successRate
• timing.averageExecutionTime, timing.minExecutionTime, timing.maxExecutionTime
• workflowPerformance.topProblematicWorkflows (sorted by failure rate)
• alerts.highFailureRate (boolean), alerts.workflowsNeedingAttention (list)

get_execution_details

Returns detailed error data for a specific workflow: error messages, failed node names and types, timestamps.

{"action": "get_execution_details", "limit": 5, "workflow_id": "<WORKFLOW_ID>"}

Health Thresholds

Notification Templates

One-Line Summary

n8n <EMOJI> | failure <RATE>% (<N> exec) | <NOTE>

Quick Status (for cron/heartbeat)

n8n monitor (<TIME> UTC) -> <EMOJI> <STATUS>

Hot spots:
<EMOJI> <WORKFLOW_NAME> - "<ERROR_MESSAGE>" (<FAILED_NODE> node)

Full Report

Combine all three endpoints for a complete picture:

1. Active workflows summary (count, categories)
2. Health status with KPIs (failure rate, timing, execution modes)
3. Hot spots table (per-workflow failure rates)
4. Error details for critical workflows
5. Alerts and recommended actions

Recommended Monitoring Patterns

Automated Cron Monitoring

• Run a monitor script every 30 minutes that calls get_workflow_executions
• Log health status to a file
• Use a heartbeat (e.g., every hour at :32) to read the log and send alerts only when status is YELLOW or RED
• Stay quiet when everything is GREEN

On-Demand Debugging

• Call get_workflow_executions to identify which workflows are failing
• Note the workflow ID
• Call get_execution_details with that ID for error patterns and failed node info

Alert Deduplication

• Hash each alert payload
• Only notify when the hash changes (new error or status change)
• Prevents notification fatigue when the same error persists

Environment

• N8N_WEBHOOK_URL: Your n8n webhook endpoint (required)
• Webhook timeout: 30 seconds (use lower limit values for large instances)

Prompts That Work Well

• "Check the health of my n8n instance"
• "Which workflows are failing?"
• "Debug errors in workflow <ID>"
• "Generate a health report for the last 100 executions"
• "What's causing failures in my data processing workflow?"

Failure Modes

• Webhook timeout: Reduce the limit parameter
• Authorization errors: Check n8n API credentials in the webhook workflow
• Empty results: Verify the webhook workflow is active and the URL is correct