anomaly-detection

Detect, classify, and respond to anomalies in time-series sensor data streams.

2|Updated Jan 31, 2026

Search Tags:#drift detection #time series #anomaly detection #statistical analysis #outlier detection #sensor data

Ready to run

npx skills add https://github.com/michaelalber/ai-toolkit --skill anomaly-detection-michaelalber

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: anomaly-detection
Source: https://github.com/michaelalber/ai-toolkit/tree/main/skills/anomaly-detection
Command: npx skills add https://github.com/michaelalber/ai-toolkit --skill anomaly-detection-michaelalber

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the critical need to identify and understand unusual patterns or deviations in sensor data streams, preventing potential failures or operational issues.

Core Features & Use Cases

Statistical Anomaly Detection: Employs multiple methods (Z-score, IQR, EWMA, CUSUM) for robust outlier detection.
Drift and Pattern Analysis: Identifies gradual drift, flatlines, and noise in time-series data.
Classification and Response: Classifies anomaly types (SPIKE, DRIFT, FLATLINE, NOISE) and suggests appropriate responses.
Use Case: Monitor temperature sensors in a data center. If a sensor shows a sudden spike or a gradual drift outside its normal operating range, this Skill will detect it, classify it as a potential overheating event, and recommend an alert or recalibration.

Quick Start

Use the anomaly-detection skill to establish a baseline for the 'temperature-sensor-1' data stream.

Frequently Asked Questions about anomaly-detection

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

FAQPage Schema

How do I detect anomalies in time-series sensor data?▼

You detect anomalies in time-series sensor data by establishing a statistical baseline and applying multiple detection methods like Z-score, IQR, EWMA, and CUSUM to identify outliers. This approach ensures consensus before flagging deviations in your data streams.

What is the best way to monitor gradual drift in sensor data streams?▼

Monitoring gradual drift in sensor data streams requires pattern analysis techniques that track sustained deviations from an established baseline. Using cumulative sum (CUSUM) and exponentially weighted moving average (EWMA) methods helps identify slow-moving trend changes accurately.

Can I classify different types of anomalies like spikes and flatlines?▼

Yes, you can classify anomalies into specific types such as SPIKE, DRIFT, FLATLINE, and NOISE. Classifying anomalies involves discriminating between actual signal changes and sensor faults to suggest appropriate alert or recalibration responses.

Do I need to establish a baseline before performing outlier detection?▼

Yes, baseline establishment is a strict prerequisite before performing outlier detection on sensor data. Accurate statistical thresholding and anomaly classification depend entirely on having a valid historical baseline to compare real-time deviations against.

How do I distinguish between a sensor fault and a legitimate signal change?▼

Distinguishing between a sensor fault and a legitimate signal change requires context-aware thresholding and multiple detection methods for consensus. This discrimination process prevents false alerts by evaluating the anomaly classification against operational context.