Skill UI

This skill enables independent auditing of AI agent, gateway, or automation workflow runs. It reconstructs events from traces, checkpoints, and tool calls to verify success claims without trusting self-reports. Users establish contracts, build claim ledgers, normalize evidence, rank witnesses, and assign verdicts like proven or contradicted. It ensures integrity in multi-step automation and separates autonomous success from human-assisted outcomes.

Design patterns for building autonomous coding agents, covering tool and function calling APIs, permission and approval systems, browser automation, and human-in-the-loop workflows inspired by Cline and OpenAI Codex.

Autonomous agents are AI systems capable of independently decomposing goals, planning actions, executing tools, and self-correcting without constant human guidance. This skill focuses on ensuring reliability in agent loops like ReAct and Plan-Execute, emphasizing goal decomposition and reflection patterns. It highlights that compounding error rates can kill autonomy, advocating for constrained, domain-specific agents over general autonomy. Build for reliability first.

This low-level SDK enables .NET developers to create and manage persistent AI agents on Azure. It supports threads, messages, runs, and tools integration. Includes features for function calling, vector store management, and file search. Ideal for building complex AI applications with C# and Azure services.

A high-level TypeScript SDK for managing Azure AI Foundry projects, enabling creation of AI agents, model deployments, datasets, and evaluations. Supports various tools like code interpreter, search, and MCP for building intelligent applications.

Build real-time voice AI applications on Azure with bidirectional WebSocket streaming using the azure-ai-voicelive Python SDK. Covers authentication, session configuration, PCM16 audio streaming, event handling and function tools.

Expert in building browser extensions for Chrome, Firefox, and cross-browser environments. Covers architecture, manifest v3, content scripts, popup UIs, monetization strategies, and Chrome Web Store publishing. Helps create secure, functional tools that users install daily.

Learn how to use a Win32 build of BusyBox to run many standard UNIX command line tools on Windows without WSL. Includes architecture detection, download instructions, and security verification.

Analyze a repository to detect its tech stack, build tools, and monorepo structure, then generate a reviewed read-only allow list for Claude Code settings.json permissions. Helps set up or audit project permission baselines safely.

Reference for building real-time voice and multimodal AI applications with Daily and Pipecat transports, covering pipelines, STT, TTS, LLM integration, tool calling, context management, and turn detection.

This skill enables the creation of comprehensive monitoring solutions for database migrations. It covers Change Data Capture (CDC) pipelines, real-time data synchronization, anomaly detection, and automated alerting. You will learn to build observability infrastructure using tools like Prometheus, Grafana, Kafka, and Debezium to track migration metrics, consumer lag, and replication status effectively.

A guide to production-ready patterns for dbt (data build tool). It covers essential practices for robust data pipeline development, including model organization (staging, intermediate, marts), implementing advanced testing strategies, ensuring data quality, and optimizing incremental processing for large datasets.