skills.rest
ruvnet
ruvnetCOMMUNITY
Explore all skills by this author →
Website𝕏@ruvGitHub

iot-anomalies

Detect and classify telemetry anomalies on Cognitum Seed IoT devices using Z-score analysis.

70.1k|8.4k|Updated Jun 2, 2025
Search Tags:#iot#anomaly detection#z-score#device monitoring#fleet health#Telemetry#firmware canary
One-click install
npx skills add https://github.com/ruvnet/claude-flow --skill iot-anomalies
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: iot-anomalies
Source: https://github.com/ruvnet/claude-flow/tree/main/plugins/ruflo-iot-cognitum/skills/iot-anomalies
Command: npx skills add https://github.com/ruvnet/claude-flow --skill iot-anomalies

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @claude-flow/plugin-iot-cognitum.

What problem does it solve?

Investigating odd device metrics, approving firmware canary rollouts, and triaging fleet-wide health alerts requires a fast, repeatable way to spot abnormal telemetry. This Skill runs Z-score anomaly detection on a device's recent telemetry and classifies what it finds.

Core Features & Use Cases

  • Z-score anomaly detection: Runs statistical detection over a device's recent telemetry via the Cognitum IoT CLI.
  • Anomaly classification: Identifies spike, flatline, drift, oscillation, pattern-break, and cluster-outlier anomaly types.
  • Quarantine recommendation: Recommends quarantining a device when the anomaly score exceeds 0.9.
  • Pattern learning: Stores detected anomaly patterns in persistent memory for future reference.
  • Use Case: Before advancing a firmware canary, run detection on a target device to confirm its telemetry is clean, or triage a fleet health alert by classifying the anomaly type on the affected device.

Quick Start

Ask the assistant to run anomaly detection on device DEVICE_ID and report the detected anomaly types and scores.

Frequently Asked Questions about iot-anomalies

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I detect anomalies in IoT device telemetry?▼

Run the Cognitum IoT anomalies command with a device ID to apply Z-score detection over the device's recent telemetry. The output classifies anomalies as spike, flatline, drift, oscillation, pattern-break, or cluster-outlier.

What anomaly types can be detected in device telemetry?▼

The detection classifies six anomaly types: spike, flatline, drift, oscillation, pattern-break, and cluster-outlier. Each type reflects a distinct deviation pattern in the telemetry time series.

When should an IoT device be quarantined based on telemetry?▼

Quarantine is recommended when the anomaly score exceeds 0.9. Scores at or below that threshold warrant review of the classified anomaly types rather than immediate isolation.

Can anomaly detection results be stored for later analysis?▼

Yes, detected anomaly patterns are persisted using the ruflo memory store under the iot-anomalies namespace, keyed by device ID with the anomaly type and score. This supports learning from past detections.

What are the limitations of Z-score anomaly detection on telemetry?▼

Z-score detection assumes roughly normal metric distributions, so highly seasonal or multimodal telemetry may produce false positives. It is best used as a triage signal combined with review of the classified anomaly type.

Related Skills

Explore More Software Engineering Skills→
lofibrainwavlofibrainwav

data-pipeline

Collect and process real-time system friction, complexity, and observer metrics.

Community
Advanced
rnavarychrnavarych
14

domain-iot:device-management

Manage IoT device lifecycles with provisioning, OTA updates, and fleet operations.

Community
Advanced
ivanshamaevivanshamaev
14

aiops-infrastructure-anomaly-detection

Detect infrastructure and data pipeline anomalies using statistical baselines and ML scoring.

Community
Advanced
ByronWilliamsCPAByronWilliamsCPA

git

Enforce Git workflow guardrails for staged files, merges, and cleanup.

Official
Advanced