syswhisper - SKILL.md Agent Skill

name: syswhisper description: "Server health monitoring with plain language Telegram alerts. Tracks CPU, memory, disk, uptime, and custom metrics. Use when: running servers/VPS instances, monitoring app health, or need simple alerting. NOT for: enterprise-scale monitoring, complex distributed systems, or regulatory compliance logging." homepage: https://pawhub.ai/syswhisper metadata: { "openpaw": { "emoji": "🔔", "requires": { "bins": ["ssh"] }, "install": [ { "id": "npm", "kind": "node", "package": "@pawhub/syswhisper", "bins": ["syswhisper"], "label": "Install SysWhisper (npm)", }, ], },

}

SysWhisper 🔔

Your servers text you when something's wrong. Plain English alerts, no cryptic error codes. Know about issues before your users do.

When to Use

✅ USE when:

Running VPS/cloud instances
Monitoring web app health
Tracking resource usage
Getting alerts for outages
Simple uptime monitoring
Disk space warnings
Process monitoring

When NOT to Use

❌ DON'T use when:

Enterprise monitoring → use Datadog, New Relic
Kubernetes clusters → use Prometheus
Compliance logging → use dedicated tools
APM tracing → use dedicated APM

Example Interactions

You: Monitor my server at 192.168.1.100

Paw: connects via SSH

✅ Server added: 192.168.1.100

Current status:

CPU: 12% (idle)
Memory: 2.1GB / 8GB (26%)
Disk: 47GB / 100GB (47%)
Uptime: 23 days

What to monitor?

High CPU (alert if >80%)
Low memory (alert if <500MB free)
Disk space (alert if <10GB free)
Uptime (alert if server restarts)

Or I can set smart defaults. Your call. 🎯

You: Smart defaults

Paw: ✅ Monitoring active

Alerts configured:

🔥 CPU >85% for 5 minutes
💾 Memory <500MB free
💿 Disk <10GB free
🔄 Server restart detected
⏰ No heartbeat for 10 minutes

I'll check every 60 seconds. You'll get Telegram pings if anything breaks.

Sleep well. 😴

3 AM alert:

Paw: 🚨 Server Alert: 192.168.1.100

Issue: High CPU usage Current: 94% (sustained for 8 minutes) Status: Critical

Top processes:

node (PID 1234) — 87% CPU
postgres (PID 5678) — 6% CPU

Actions:

/sys kill 1234 (kill node process)
/sys restart node-app (restart service)
/sys ssh (get SSH command)

What do you want to do? 🔧

You: /sys restart node-app

Paw: Restarting node-app on 192.168.1.100...

✅ Service restarted

Old PID: 1234 (killed)
New PID: 9876 (running)
CPU: 12% (back to normal)

Status: Resolved ✅

I'll keep watching. Go back to sleep. 😴

Commands

# Add server
syswhisper add 192.168.1.100 --name "web-server"

# Configure alerts
syswhisper alert cpu --threshold 85 --duration 5m

# Show server status
syswhisper status

# View alert history
syswhisper alerts history

# Test alert
syswhisper alerts test

Telegram Integration

Quick status:

You: /sys status

Paw: All servers: 3 online

✅ web-server — healthy ✅ db-server — healthy ⚠️ worker-1 — high memory (85%)

React to worker-1 for details.

Pricing

Free tier: 1 server, basic alerts
Pro: $8/month — 10 servers, custom alerts

Install: pawhub.ai/syswhisper

Sleep soundly. Paw's watching your servers. 🔔🐾