Can I use this for streaming data applications?

Yes, specific backends like ECOD, HBOS, and LODA are highly optimized for speed and are suitable for streaming-friendly use cases.

What algorithms are supported for anomaly detection?

The skill supports over 12 backends including Isolation Forest (iforest), ECOD, LOF, kNN, OCSVM, COPOD, HBOS, and deep learning Autoencoders.

How do I choose the right backend for my data?

The skill provides a decision tree: use iforest for general tabular data, ECOD for high-dimensional data, and Autoencoders for complex non-linear patterns.

Does the skill handle data preprocessing?

Yes, it includes built-in normalization methods such as standardization and z-score scaling to ensure data is properly prepared for the detection models.

LlamaFarm Anomaly Detection

Name: LlamaFarm Anomaly Detection
Author: llama-farm

byllama-farm

0•

Data Science & ML

Detects outliers and anomalies in datasets using over 12 machine learning backends and automated training workflows.

This skill empowers Claude to perform sophisticated anomaly detection tasks within the LlamaFarm ecosystem. It provides a robust interface for training models, scoring data points, and identifying outliers across a wide variety of algorithms including Isolation Forests, ECOD, and deep learning Autoencoders. Whether dealing with high-dimensional data, streaming sources, or complex non-linear patterns, this skill guides users through selecting the right backend, preparing data with built-in normalization, and managing the full model lifecycle from training to deployment.

Key Features

010 GitHub stars

02Support for 12+ machine learning backends like Isolation Forest, ECOD, and Autoencoders

03Built-in data normalization methods including standardization and Z-score scaling

04Integrated decision-tree guidance for selecting algorithms based on speed and accuracy

05Automated model training, scoring, and outlier detection via REST API

06Model lifecycle management for saving, loading, and versioning trained models

Use Cases

01Monitoring system logs and metrics to detect infrastructure performance issues

02Identifying fraudulent transactions or suspicious activity in financial datasets

03Finding defective units in manufacturing sensor data or quality control processes

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add llama-farm/claude-code-marketplace anomaly-detection

For use in Claude.ai and ChatGPT

Download Skill