Back to Tutorials

Sending Notifications from AI

Intermediate
20 minutes
What You'll Learn

This tutorial will teach you how to set up and use the notification system in Plugged.in, allowing your AI assistant to send real-time alerts and updates.

How to enable notifications in your profile
Using the pluggedin_send_notification tool
Setting up email notifications for critical alerts
Best practices for notification severity levels
Prerequisites
  • Plugged.in account with MCP proxy configured
  • MCP client (Claude Desktop, Cursor, etc.) connected
  • Email configured in settings (optional, for email notifications)

1
Enable Notifications

First, ensure notifications are enabled for your current profile. This allows the MCP proxy to log activities and send notifications.

Screenshot: Profile settings with notifications toggle

Shows the notification settings in your profile

2
Configure Email Notifications (Optional)

For critical alerts, you can enable email delivery. This ensures you receive important notifications even when not actively using Plugged.in.

Email Setup Steps
  1. Go to Settings → Notifications
  2. Toggle 'Email Notifications' on
  3. Verify your email address if not already done

Screenshot: Email notification settings

Shows email configuration options

3
Using the Notification Tool

The pluggedin_send_notification tool is available in your MCP client. Here's how to use it effectively:

Simple Notification

Send a basic notification that appears in the Plugged.in notification center:

Use pluggedin_send_notification to send "Task completed successfully!"

Screenshot: Notification appearing in the bell menu

Shows how the notification appears in the UI

4
Practical Examples

Deployment Notifications

Notify team members about deployment status:

1. Query RAG for deployment checklist
2. Execute deployment steps
3. Send notification "Deployment to production complete" with SUCCESS severity
4. If errors occur, send ALERT with email enabled
Analysis Results

Send notifications based on analysis outcomes:

After analyzing logs:
- If errors found: Send "Found 3 critical errors in logs" with ALERT
- If warnings: Send "5 warnings detected" with WARNING
- If clean: Send "Log analysis complete - no issues" with SUCCESS
System Monitoring

Alert on system resource usage:

Monitor system resources and:
- CPU > 80%: Send "High CPU usage: 85%" with WARNING
- Memory > 90%: Send "Critical memory usage: 92%" with ALERT + email
- Disk full: Send "Disk space critical: 2GB remaining" with ALERT + email

Best Practices

Keep it Meaningful

Include specific details like counts, percentages, or status codes

Use Appropriate Severity

Reserve ALERT for critical issues that need immediate attention

Email Sparingly

Only enable email for notifications that require action when away from the app

Add Context

Include timestamp, affected systems, or relevant identifiers

Next Steps

Enhance your automation workflows with these related tutorials: