Skill UI

This skill outlines the development of an automated pipeline to ingest raw Indicators of Compromise (IOCs) from various sources. It performs normalization, defanging (e.g., converting 'http://' to 'hxxp://'), and converts indicators into the standardized STIX 2.1 format. The processed, safe indicators can then be securely distributed via TAXII feeds, MISP, and other threat intelligence platforms.

A comprehensive guide and methodology for proactively hunting adversary abuse of legitimate cloud services (Azure, AWS, GCP, SaaS). This technique focuses on identifying Command and Control (C2), data staging, and exfiltration methods that leverage normal cloud functionality, making detection challenging.

This advanced workflow orchestrates the entire research idea lifecycle. Starting from a broad research direction, it systematically executes literature mapping, creative brainstorming, novelty verification, and proposal refinement. The final output is a validated IDEA_REPORT, a detailed proposal, and a concrete experiment plan, transforming preliminary concepts into actionable, structured research projects. Ideal for users needing a full, end-to-end research guidance system.

This skill guides the implementation of network traffic baselining using historical NetFlow/IPFIX data. It leverages Python's pandas library to compute detailed statistical profiles, including hourly/daily distributions, per-host metrics, and protocol ratios. Anomalies are robustly detected using z-score and Interquartile Range (IQR) methods, assisting SOC analysts in identifying unusual activities like data exfiltration or beaconing.

Tetragon, a CNCF project under Cilium, provides advanced, kernel-level runtime security observability and policy enforcement in Kubernetes clusters. By utilizing eBPF, it monitors process execution, file access, and network connections with minimal overhead, enabling detection and prevention of container escapes, unauthorized access, and suspicious activity.

A specialized tool for formalizing and verifying complex mathematical proofs, particularly in Machine Learning and AI theory. It guides users through structuring theorems, identifying explicit assumptions, performing feasibility triage, and ensuring every logical step is rigorously justified, preventing the fabrication of proofs.

A powerful tool for conducting comprehensive academic research and literature reviews. It searches, analyzes, and synthesizes key ideas from multiple sources, including local PDF libraries (Zotero, Obsidian), the general web, and specialized academic databases (Semantic Scholar, DeepXiv). Use this when you need to map the historical context, identify related works, or understand complex concepts across multiple studies on a given topic.

This skill automates the entire academic research lifecycle. It orchestrates stages from broad idea discovery and literature review to experimental implementation, rigorous cross-model review, and finally, generating a polished research paper PDF. Ideal for researchers needing an end-to-end, autonomous workflow for academic publication preparation.

Get a deep, multi-round, and adversarial critical review of your research work, ideas, or drafts. This skill simulates the perspective of a senior ML reviewer (e.g., NeurIPS/ICML level) using dedicated external backends (Codex or Manual). It is designed to find logical gaps, missing experiments, and narrative weaknesses, providing brutal honesty to elevate your academic contribution.

This module provides a centralized interface for the AgentHub message board. It allows multiple AI agents and the coordinator to communicate, track tasks, and log results through structured markdown posts across dedicated channels (Dispatch, Progress, Results). It is essential for monitoring complex multi-agent workflows and ensuring transparent collaboration and auditing.

Soft Predict Future is a structured foresight engine that triggers whenever the user asks about future outcomes or scenarios. It enforces a nine-step analytical pipeline (validation, signal collection, scoring, driver extraction, etc.), highlights required searches and inputs, and directs the user to web-enabled reasoning for predictive analysis.

This tool audits the token consumption across all components (agents, skills, rules, MCP servers) within a Claude Code session. It analyzes token overhead, identifies bloat (e.g., overly long descriptions, redundant components), and generates a detailed context budget report. Use it to maintain optimal context window usage, improve performance, and prevent context saturation.