Skill UI

Provides Python access to openFDA endpoints for drugs, devices, foods, veterinary products, and substances, enabling regulatory analysis, safety monitoring, recalls tracking, and pharmacovigilance research.

A specialized skill for drafting high-impact, highly personalized investor communications. It covers cold emails, warm intro requests, follow-up updates, and overall pipeline management. The skill ensures every message is concrete, actionable, and structured to maximize engagement and minimize friction with busy investors.

AgentMail provides a comprehensive REST API email infrastructure specifically designed for AI agents. Agents can provision dedicated inboxes, send and receive emails, handle external service verifications (e.g., GitHub signups), and manage real-time inbound webhooks. The system includes a 'karma' mechanism to ensure high domain reputation and prevent abuse, making it reliable for complex automation workflows.

Parses Cobalt Strike malleable C2 profiles via pyMalleableC2 to pull beacon configs, HTTP communication patterns, sleep/jitter settings, and combines JARM scans for C2 detection when hunting suspected infrastructure or building IDS signatures.

This guide provides comprehensive steps to harden Docker containers and associated runtime environments based on the CIS Docker Benchmark v1.8.0. It covers implementing least-privilege principles, such as running as non-root users, dropping capabilities, using read-only filesystems, and applying seccomp/AppArmor profiles. Use this when preparing containers for production deployment or conducting security audits to minimize the attack surface.

This skill provides advanced techniques and code examples for monitoring, mapping, and pivoting across adversary-controlled infrastructure. It leverages Passive DNS, Certificate Transparency logs, Shodan/Censys scanning, and network fingerprinting (JARM/JA3S) to discover C2 servers, phishing domains, and exploitable assets over time, which is crucial for advanced security operations and threat intelligence gathering.

This guide outlines the implementation of Remote Browser Isolation (RBI) as a core component of a Zero Trust architecture. It details policies for URL risk classification, Data Loss Prevention (DLP) controls, and Content Disarm and Reconstruction (CDR). This is essential for hardening web access, protecting against zero-day exploits, phishing, and preventing data exfiltration from untrusted web content.

A comprehensive tool designed to reorganize and optimize your professional networks on X and LinkedIn. Instead of treating outbound connections as mere lead lists, it analyzes your social graph by identifying pruning opportunities, recommending high-value connections to keep, and generating warm outreach drafts. Use it to enhance relationship quality and strategic growth.

An advanced tool for network intelligence that calculates weighted graph scores to map the optimal path to a target. It identifies potential introducers, measures 'bridge value' across connections, and helps determine whether a connection requires a warm introduction or direct outreach on platforms like X and LinkedIn. Ideal for users needing the raw ranking methodology.

This skill manages comprehensive software development life cycle operations—including issue triage, PR reviews, version releases, bug fixing, and marketing content generation. It operates using a parallel "agent army" of dynamic agents (10-20 per task), requiring the user to adopt an Engineering Manager (EM) mindset. Key principles involve rigorous citation checking (anti-hallucination) and applying multiple roles (Product Owner, Open Source Contributor, Architect) simultaneously to ensure robust, safe, and user-centric outcomes across all platforms.

Provides the Google Model Armor CLI to filter user-generated content and run template-guided sanitization for prompts or responses, helping teams enforce safety policies during automation workflows and schema-driven inspections.

Creates a new Google Model Armor template via `gws modelarmor +create-template`, requiring project, location, and template ID, with optional preset or JSON payload; aimed at bootstrapping safety filters before sanitization commands.