Skill UI

A comprehensive guide for implementing spatial, mixed-reality user interfaces. Inspired by Apple Vision Pro and AR aesthetics, this pattern uses environmental transparency, blurred glass-like panels, and dynamic lighting effects to create deeply integrated, context-aware layouts for web and mobile applications.

This skill enables querying Delta Lake tables and raw files stored in OneLake using DuckDB. It supports two primary modes: local execution via CLI/Python for ad-hoc analysis, and in-notebook execution within Fabric, combining DuckDB's speed with notebook context. Ideal for data exploration, schema discovery, and data quality validation across the Fabric ecosystem.

This skill scaffolds the necessary configurations for advanced engineering workflows. It guides the user through setting up the issue tracking system (GitHub, GitLab, or local markdown), defining the canonical triage label vocabulary, and establishing the domain documentation context (single or multi-context). It must be run once to initialize the repository for subsequent agent-driven development skills.

This skill guides the systematic triage of project issues and external pull requests (PRs). It moves development items through a defined state machine using roles (e.g., bug, enhancement) and states (e.g., ready-for-agent, wontfix). The process involves context gathering, cross-codebase verification, deep questioning (grilling), and generating comprehensive, agent-ready briefs or marking them for human implementation.

This document provides a comprehensive guide to the principles and vocabulary needed to write high-quality, predictable, and effective skills. It covers advanced concepts like model-invoked vs. user-invoked skills, managing context load, structuring information hierarchies (steps vs. reference), and implementing progressive disclosure to improve agent predictability.

Provides a comprehensive guide to implementing Zero Trust Network Architecture on Google Cloud Platform (GCP) using Identity-Aware Proxy (IAP). It enforces rigorous, per-request identity verification and context-aware authorization for services like Compute Engine, App Engine, and Cloud Run. This ensures that only authenticated users who meet specific criteria (e.g., device posture, corporate location, time window) can access protected resources, eliminating reliance on public IPs or VPNs.

This skill provides comprehensive security auditing for Model Context Protocol (MCP) servers and their associated tool metadata. It is designed to detect sophisticated supply chain attacks, including tool poisoning (indirect prompt injection), Server-Side Request Forgery (SSRF), tool shadowing, and unauthenticated exposure. It is essential for securing AI agent stacks before integrating new tools or during CI/CD processes.

A comprehensive red-teaming skill for assessing the security posture of Retrieval-Augmented Generation (RAG) pipelines. It systematically probes two critical injection surfaces: poisoned retrieved context (indirect prompt injection) and embedding manipulation. Use this skill to validate retrieval guardrails, ensure data leak prevention, and demonstrate vulnerabilities in LLM-powered knowledge assistants.