AI-Powered Server Monitoring
Quick Install BETA
$ # One-liner installation
curl -fsSL spyd.sh | sh
⚠️ Currently in beta. Best used on test servers.
[DIAGNOSE]

Senior SysAdmin Intelligence

AI trained to think like a battle-hardened sysadmin. Identifies root cause, not just symptoms.

[DETECT]

Attack Pattern Recognition

Crypto miners, DDoS, brute force, fork bombs — detected and explained in plain English.

[ALERT]

Incident Correlation

Groups related alerts into incidents. One notification, not alert fatigue.

How It Works

[▓▓▓░]

Collect

CPU, memory, disk, load, top processes every 60 seconds

→
<AI/>

Analyze

Senior SysAdmin methodology: Observe → Correlate → Identify

→
[###]

Correlate

Group related anomalies into incidents, suppress duplicates

→
[!]

Alert

Plain English diagnosis with real PIDs and safe-first commands

Quick Start

terminal
$ curl -fsSL spyd.sh | sh
✓ Spyd installed successfully!
$ spyd init
✓ Configuration created
$ spyd start
✓ Daemon started (PID: 12345)
$ spyd status
● Daemon: Running CPU: 45.2% Memory: 62.1% Disk: 55.3%
$ █
spyd init Initialize configuration
spyd start Start daemon
spyd status Show current metrics
spyd watch Live dashboard
spyd alerts View alert history
spyd logs View daemon logs

Real AI Diagnosis

spyd alert (Telegram)
🚨 CPU usage 100.0% exceeds threshold 80.0%
Summary: Your server is being hammered by automated scripts. Multiple bash processes are consuming all CPU — this looks like a stress test or denial-of-service attempt.
Root Cause: Fork bomb pattern detected. PIDs 3685587, 3685588, 3685589, 3685590 spawned from same parent.
Suggested Fix:
$ kill -15 3685587 3685588 3685589 3685590
Confidence: 90% • Mode: hybrid • Severity: critical

Why Spyd?

λ

AI-Powered

Plain-English explanations, not cryptic errors

#

Privacy-First

Your data, your API keys. Zero telemetry

~/

Lightweight

Single binary, <50MB RAM, <5% CPU

$0

Free Forever

Open source (MIT License)

>>>

Multi-Channel

Email, Slack, Telegram, Discord, Desktop (ntfy), Webhooks

&&

Smart Alerts

Incident correlation, no notification spam

Works With Your AI

[CLOUD]

OpenAI / Anthropic

GPT-4o-mini, Claude — fast, accurate, pay-per-use

[LOCAL]

Ollama / LiteLLM

Run Llama, Mistral locally. 100% private, zero cost

[ENTERPRISE]

Azure OpenAI

Your Azure deployment, your compliance requirements