Skill UI

A comprehensive guide for security teams on detecting and responding to unauthorized crypto mining operations in cloud environments (AWS, Azure). It details a multi-layered approach, utilizing cost anomaly detection, compute utilization monitoring, GuardDuty findings, and network flow log analysis (VPC/KQL) to identify resource hijacking attempts and suspicious activity across EC2, ECS, and EKS workloads.

A specialized tool for detecting anomalies within DNP3 (Distributed Network Protocol 3) communications used in critical SCADA and energy sector infrastructure. It employs deep packet inspection and baseline comparison to monitor for unauthorized control commands, protocol violations, suspicious firmware update attempts, and deviations from established operational traffic patterns.

This skill provides structured procedures and Python examples for detecting internal data exfiltration. It analyzes endpoint, cloud, and email logs by establishing behavioral baselines and performing statistical anomaly detection. Use cases include investigating insider threats, monitoring DLP policy violations, and building robust User and Entity Behavior Analytics (UEBA) rules for SOC teams.

This skill implements User and Entity Behavior Analytics (UEBA) to detect sophisticated insider threats. It builds behavioral baselines from diverse log sources (authentication, file access, network) using Elasticsearch/OpenSearch. It calculates anomaly scores and performs peer group analysis to identify deviations indicative of data exfiltration, privilege abuse, or unauthorized access patterns.

This skill provides a comprehensive framework for detecting Living Off the Land Binaries (LOLBAS) abuse, such as misuse of certutil, regsvr32, and mshta. It leverages process telemetry from Sysmon and Windows Event Logs, combined with advanced Sigma rule-based detection and parent-child process anomaly analysis. Ideal for SOC analysts and threat hunters investigating sophisticated adversaries aiming to evade traditional security controls.

This skill provides comprehensive anomaly detection for Modbus/TCP and Modbus RTU traffic in Operational Technology (OT) and Industrial Control Systems (ICS). It monitors function codes, validates register access, analyzes timing, and detects unauthorized clients. By utilizing advanced tools like Zeek, Suricata, and custom Python models, it helps establish behavioral baselines and identify potential cyber threats.

This skill provides a comprehensive guide to deploying Zeek (formerly Bro) for passive network security monitoring. It details how to generate structured logs (DNS, HTTP, SSL, etc.) and write custom Zeek scripts to detect sophisticated anomalies like DNS tunneling and beaconing. Ideal for threat hunting, security incident response, and complementing traditional IDS solutions.

This skill provides comprehensive guidance on detecting network reconnaissance and port scanning activities using industry-standard IDS/IPS systems like Suricata and Snort. It covers various scanning techniques (e.g., SYN, FIN, NULL, Xmas scans, UDP sweeps) and advanced detection methods, including signature matching, threshold-based alerting, and traffic anomaly analysis to identify tools like Nmap and Masscan.

This skill detects malicious Command and Control (C2) communications tunneled through the DNS protocol. It analyzes DNS query logs for indicators such as DNS tunneling tools (Iodine, dnscat2), Domain Generation Algorithms (DGA), and encoded payloads within TXT/CNAME records. Techniques include Shannon entropy analysis of subdomains, statistical anomaly detection, and machine learning classification. Ideal for network security analysts conducting deep threat hunting and forensic investigations.

Provides methodology and scripts for proactive threat hunting, IOC sweeps, and statistical anomaly detection that surface attacker activity before automated alerts fire.

Aeon is a comprehensive, scikit-learn compatible Python toolkit designed for advanced time series analysis. It supports a wide range of temporal machine learning tasks, including classification, regression, clustering, forecasting, anomaly detection, and similarity search. Use this skill when working with sequential patterns, time-indexed observations, or complex temporal data that requires specialized algorithms.

This comprehensive benchmark suite evaluates the performance of Vision-Language Models (VLM) on complex video anomaly detection tasks using the SmartHome-Bench dataset. It covers a wide range of real-world smart home scenarios, including security breaches, elderly fall detection, and unusual pet behavior. Unlike single-frame analysis, this benchmark requires deep multi-frame video understanding, making it a robust tool for testing the advanced AI capabilities of video monitoring systems.