Skill UI

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.

This skill creates and synchronizes prompt-based AI agents directly into Azure AI Foundry using the Agent Service REST API. By defining agents in a local JSON manifest, the skill registers them server-side, making them instantly available for evaluation and invocation within the Foundry ecosystem. It is essential for deploying, updating, or managing AI agents at scale within the Azure cloud environment.

Run a GEPA learning cycle via metaharness learn to optimize a harness genome against a SWE-bench-style slice manifest. Supports dry-run for cost estimation and --run for explicit execution. Requires a metaharness repo checkout.

This guide outlines the stringent standards for ensuring the reproducibility of Natural Language Processing (NLP) research. It teaches authors how to treat the Responsible NLP checklist as a binding contract by meticulously 'pinning' every variable—including model identifiers, API access dates, prompt strings, random seeds, and evaluation metrics. It details creating a comprehensive reproducibility manifest to make research results verifiable and defensible.

OpenSquilla-compatible adapter that generates browser-playable AI audio such as voiceovers, narration, music, and sound effects through OpenRouter audio models. It streams OpenAI gpt-audio output, writes WAV assets, and reports structured AUDIO_READY manifests for webpage integration.

An adapter for generating short, browser-playable video assets using OpenRouter's asynchronous video generation API. This skill is designed for meta-skills to integrate dynamic multimedia content into web pages. It handles the full job lifecycle—submitting the request, polling for completion, and downloading the final MP4 asset—and adheres strictly to the required manifest reporting format.

This document outlines the canonical contract for deploying model inference services on Itô compute. It details the secure handoff process of an immutable serving manifest after a compute booking is verified. The process enforces multiple strict gates, including entitlement checks, manifest integrity validation, idempotency key usage, and final health checks, ensuring robust and reliable deployment of advanced AI models.

This skill orchestrates advanced Machine Learning training workflows on dedicated compute clusters. It manages complex stages (pre-training, fine-tuning, RL) by ensuring strict adherence to immutable manifests, fresh booking entitlements, and multi-step confirmation gates. It acts as a canonical interface for job submission, emphasizing security, idempotency, and resource validation before execution.