azure-ai

Implements Azure AI Search, Speech, OpenAI, and Document Intelligence integrations via MCP tools and SDK references.

Updated Jan 23, 2026
One-click install
npx skills add https://github.com/stephschofield/beth --skill azure-ai-stephschofield
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: azure-ai
Source: https://github.com/stephschofield/beth/tree/main/.github/skills/azure-ai
Command: npx skills add https://github.com/stephschofield/beth --skill azure-ai-stephschofield

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Working with Azure AI services requires knowing which service fits a task, which SDK patterns apply per language, and how to authenticate correctly. This Skill consolidates guidance for Azure AI Search, Speech, OpenAI, and Document Intelligence so you can implement search, transcription, OCR, and moderation without hunting through documentation. ## Core Features & Use Cases - Service Selection Guidance: Maps tasks like vector search, speech-to-text, and OCR to the right Azure service, MCP tool, or CLI command. - Multi-Language SDK References: Provides condensed quick-start patterns and best practices for Python, TypeScript, .NET, and Java SDKs covering Search, OpenAI, Vision, Translation, Content Safety, and Document Intelligence. - Authentication Best Practices: Explains when to use DefaultAzureCredential versus ManagedIdentityCredential with environment-aware code samples. - Use Case: When asked to add hybrid vector search to an app, the Skill directs you to the azure__search MCP tools or the azure-search-documents SDK reference with batching and semantic ranking guidance. ## Quick Start Ask the agent to set up hybrid vector search on an Azure AI Search index using the TypeScript SDK.

Frequently Asked Questions about azure-ai

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I implement hybrid vector search with Azure AI Search?▼

Hybrid search combines vector and keyword queries in one request. In Python, use VectorizedQuery with SearchClient; in TypeScript, pass vectorSearchOptions.queries with kind "vector". Enable semantic ranking for natural language queries to improve relevance.

How do I transcribe audio with Azure AI Speech in Python?▼

Install azure-ai-transcription and create a TranscriptionClient with your endpoint and subscription key. Use begin_transcription for batch files in blob storage with diarization, or begin_stream_transcription for real-time audio.

Should I use DefaultAzureCredential in production on Azure?▼

No. DefaultAzureCredential is intended for local development because its fallback chain adds latency and unpredictability. In production, use ManagedIdentityCredential with system- or user-assigned managed identities and Azure RBAC.

Does Azure Document Intelligence support TypeScript?▼

Yes. Install @azure-rest/ai-document-intelligence and call the /documentModels/{modelId}:analyze path. Analysis is a long-running operation, so use getLongRunningPoller and pollUntilDone, and check isUnexpected for error handling.

What MCP tools are available for Azure AI Search and Speech?▼

When the Azure MCP server is enabled, azure__search supports search_index_list, search_index_get, and search_query commands, while azure__speech supports speech_transcribe and speech_synthesize. If MCP is not enabled, run /azure:setup or enable it via /mcp.