Skill UI

An expert guide for deploying and optimizing Large Language Models (LLMs) entirely on local hardware. Mastering the entire local AI ecosystem, including inference engines like Ollama, vLLM, and llama.cpp. Provides deep knowledge on quantization formats (GGUF, EXL2, AWQ), VRAM calculation, and ensuring privacy-first, offline deployment for state-of-the-art models.

This skill automates the end-to-end workflow for deploying and executing ML training and inference experiments. It intelligently detects the target environment (Local GPU, Remote SSH, Vast.ai, Modal serverless) and handles necessary steps like GPU pre-checks, code synchronization (rsync/git), integrating ML tracking tools (W&B), and launching the compute job via screen sessions or serverless deployment. Use when initiating a complex training job.

Guides fine-tuning and evaluation of OpenVLA-OFT/OFT+ for robot action generation, covering LoRA adaptation, continuous action heads, FiLM conditioning, and LIBERO/ALOHA benchmarks plus deployment considerations.

Assesses AI/ML systems for prompt injection, jailbreak, model inversion, data poisoning, and agent tool abuse using MITRE ATLAS mapping, signature detection, and scoring metrics before deployment.

A comprehensive guide for developers to build advanced, declarative agents for Microsoft 365 Copilot. This skill details the integration process using a Model Context Protocol (MCP) server, allowing the agent to connect with external systems, select tools, and handle complex authentication (OAuth 2.0/SSO). You will learn to generate key JSON manifests required for deployment.

Guides hooking Phoenix AI services with OpenInference tracing, covering setup, spans, instrumentation, and production deployment for LLM workflows using Python or TypeScript.

This tool generates a complete, structured TypeSpec declarative agent definition tailored for Microsoft 365 Copilot. It systematically creates necessary components including detailed role instructions, conversation starters, and a comprehensive list of required capabilities (e.g., WebSearch, OneDrive, Teams, CodeInterpreter). This ensures the agent is fully functional and ready for deployment within the M365 ecosystem.

This guide provides comprehensive strategies to significantly reduce spending on the Anthropic Claude API. Key techniques include model routing (selecting the most cost-effective model for the task), prompt caching for repeated prompts, using Message Batches for asynchronous tasks, and implementing real-time spend tracking and budget alerts. Essential for managing large-scale, high-cost AI deployments and optimizing cloud LLM expenditures.

A comprehensive readiness checklist for deploying clinical AI systems, such as Abridge, within regulated healthcare environments. It covers critical aspects including HIPAA compliance, Business Associate Agreements (BAA), EHR integration validation (FHIR R4), robust infrastructure security (TLS 1.3, secrets management), and provides structured scripts for pre-launch validation and post-launch rollback planning.

This module provides comprehensive observability for Anthropic's Claude API integrations. It enables the tracking of critical operational metrics including latency, token usage (input/output), estimated API costs, and error rates. Use it to set up structured logging and Prometheus metrics for robust monitoring and alerting, ensuring reliable and cost-effective AI application deployments.

A comprehensive checklist for deploying Claude API integrations to a production environment. It covers essential best practices for reliability, observability, and cost control, including authentication procedures, advanced error handling, rate limit management, health check implementation, and infrastructure deployment strategies (e.g., canary releases) to ensure a smooth and stable launch.

Provides strategies to optimize costs when utilizing the Cohere API. This guide covers model tier selection (economy, standard, premium), implementing token budgets, reducing embedding storage costs (e.g., using int8), and establishing comprehensive usage trackers. It is essential for maintaining cost predictability in production AI deployments.