Skill UI

This skill provides a comprehensive framework for security auditing LLM agents that possess external tools (file system access, command execution, network calls). It employs advanced dialogue probing techniques—such as command injection, path traversal, and SSRF attempts—to identify vulnerabilities where the agent might be induced to execute unexpected or malicious code, or access sensitive resources. Essential for red teaming and securing agent functionalities.

A comprehensive playbook detailing expert techniques for OS command injection vulnerabilities. It covers all shell metacharacters, blind injection, time-based detection, out-of-band exfiltration methods, and common vulnerable code patterns across PHP, Python, and Node.js. Essential for security analysts and penetration testers.

This expert playbook provides comprehensive techniques for exploiting Server-Side Template Injection (SSTI). It covers polyglot detection probes, engine fingerprinting (Jinja2, Twig, FreeMarker, etc.), sandbox escape mechanisms (MRO chains), and advanced bypass techniques. Essential reading for security researchers and penetration testers assessing application vulnerabilities.

This playbook details the methodology for detecting and exploiting subdomain takeover vulnerabilities. It focuses on identifying CNAME/NS/MX records pointing to deprovisioned or unclaimed cloud resources (e.g., AWS S3 buckets, Heroku apps). Understanding these weaknesses allows attackers to serve malicious content under the victim's trusted domain, leading to phishing, credential theft, and security bypasses.

A risk-paranoid interrogation tool designed to proactively vet any technical plan that handles sensitive data, affects compliance scope, or touches production access. It forces a comprehensive review across threat modeling, blast radius quantification, detection capabilities, incident response planning, regulatory compliance, and vendor security.

This skill provides a comprehensive solution for monitoring brand impersonation attacks across multiple digital vectors. It systematically scans domains, social media platforms, mobile app stores, and the dark web to detect phishing campaigns, typosquatting, and unauthorized brand usage. It is essential for security teams conducting proactive threat intelligence gathering and vulnerability assessments to safeguard brand integrity.

This comprehensive guide details how to use industry-leading tools like TruffleHog and git-secrets to scan source code repositories, full git history, and configuration files for hardcoded AWS credentials and secrets. It is essential for integrating secrets detection into CI/CD pipelines, performing proactive security audits, and ensuring compliance by preventing sensitive keys from reaching production environments.

Stratus Red Team is an open-source tool for detection engineering and purple teaming. It atomistically detonates granular, MITRE ATT&CK-mapped offensive techniques across AWS, Azure, GCP, and Kubernetes. The tool executes a full, repeatable lifecycle (warmup, detonate, revert, cleanup) to generate realistic attacker telemetry. Use it to validate if cloud security tools (like CloudTrail or Sentinel) reliably trigger on simulated attacks, or to safely practice red teaming.

Exploits privileged pods, host mounts, runC CVEs, and exposed Docker sockets to break out of a container and reach the underlying host during an authorized container-security assessment. Use when executing an approved breakout test, demonstrating the real impact of a privileged or hostPath workload, or validating that escape mitigations actually hold. Keywords: container breakout, privileged, hostPath, docker.sock, runC CVE-2019-5736, CVE-2024-21626, release_agent, nsenter. Do not use for defensive detection of these techniques - use detecting-container-escape-attempts.

A comprehensive audit worker that integrates Open Infrastructure Architecture (OIA) manifest analysis, threat modeling, and MCP security scanning. It computes a composite worst-case severity score and persists time-stamped audit records in a memory namespace. Primarily designed for scheduled CI/CD pipelines to detect architectural drift and security regressions over time.

Runs the upstream metaharness-darwin security bench, evolving a champion security-detection harness against a 10-vuln/9-decoy corpus. Grades TPR/FPR/patch-pass/repro/unsafe vs four baselines, supporting ADR-155 nightly self-learning security harness. Degrades gracefully when the upstream package is absent.

A comprehensive guide detailing the high standards required for presenting security evidence at major conferences. It covers methodologies for proving claims—including attacks, defenses, and detection systems—across live internet measurements, controlled testbeds, and emulation. Key requirements include adaptive adversary modeling, rigorous measurement hygiene, and adherence to ethical review protocols to ensure results are reproducible and relevant to real-world networks.