A field guide to time-series anomaly detection tools

What it does

This repository is a curated alphabetical index of open-source software and datasets for anomaly detection on time-series data. It collects libraries, frameworks, and analysis tools in sortable tables and labels each project as maintained or not maintained based on commit activity or author disclosure. The scope covers dedicated anomaly detectors, forecasting libraries, and general time-series analysis packages.

The interesting bit

The real value is the maintenance flag: a simple checkmark or cross tells you whether the latest commit is less than a year old, sparing you the archaeology of dead issues and broken dependencies. That binary signal is crude—stable libraries can look abandoned—but it is surprisingly rare in awesome lists.

Key highlights

• Covers roughly three dozen anomaly detection tools across Python, Java, R, Go, C++, and Julia.
• Splits entries into core anomaly detection and related software such as forecasting and time-series analysis.
• Maintenance status is defined transparently: latest commit older than one year or explicit author note equals “not maintained.”
• Includes high-profile entries like Twitter’s AnomalyDetection, Netflix’s Surus, and NASA’s Telemanom, many now flagged as unmaintained.
• Provides a citation policy for researchers who want to reference the collection.

Caveats

• The maintenance badge is strictly commit-age-based, so mature but stable libraries may be marked as unmaintained unfairly.
• The README is truncated in the provided source, so the datasets section and any additional categories are not visible here.

Verdict

Worth bookmarking if you are evaluating tooling for time-series outlier detection and want to avoid abandoned repositories. Skip it if you are looking for a single installable framework or a hands-on tutorial.