azure-ai

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

Updated May 12, 2026
One-click install
npx skills add https://github.com/sohamda/apex-try-out-demo-repo --skill azure-ai-sohamda
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: azure-ai
Source: https://github.com/sohamda/apex-try-out-demo-repo/tree/main/.archive/_archived_skills/azure-ai
Command: npx skills add https://github.com/sohamda/apex-try-out-demo-repo --skill azure-ai-sohamda

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Developers integrating Azure AI services often struggle to choose the right service, SDK, and authentication pattern for tasks like search, transcription, OCR, and content moderation. This Skill provides a consolidated decision guide with MCP tool commands, SDK quick references, and authentication best practices. ## Core Features & Use Cases - Service Selection Guidance: Maps tasks like vector search, speech-to-text, and form extraction to the correct Azure service (AI Search, Speech, OpenAI, Document Intelligence) with MCP tool commands. - Multi-Language SDK References: Condensed quick-start patterns for Python, TypeScript, .NET, and Java covering Search, OpenAI, Vision, Translation, Transcription, Content Safety, and Document Intelligence. - Authentication Best Practices: Environment-aware credential patterns using ManagedIdentityCredential in production and DefaultAzureCredential for local development. - Use Case: When building a RAG application, use this Skill to configure hybrid vector search with azure__search MCP commands, then follow the Python SDK reference to implement batch document uploads with SearchIndexingBufferedSender. ## Quick Start Ask the agent to list Azure AI Search indexes and run a hybrid vector query using the azure-ai skill with Azure MCP enabled.

Frequently Asked Questions about azure-ai

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

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

Use the azure__search MCP tool with search_query command, or the SDK with VectorizedQuery for Python and vectorSearchOptions for TypeScript. Combine vector and keyword queries for hybrid search, and enable semantic ranking for natural language queries.

How to transcribe audio to text with Azure Speech?▼

Use the azure__speech MCP tool with the speech_transcribe command, or the Python azure-ai-transcription SDK. Batch transcription supports diarization for multiple speakers, while real-time streaming handles live audio input.

Should I use DefaultAzureCredential in production on Azure?▼

No, use ManagedIdentityCredential in production for deterministic, secret-free authentication. DefaultAzureCredential is intended for local development only because its fallback chain adds latency and unpredictable behavior across environments.

Does Azure Document Intelligence support custom form models?▼

Yes, use DocumentIntelligenceAdministrationClient to build custom models with BuildDocumentModelOptions in template or neural mode. Prebuilt models like prebuilt-invoice handle common documents, while custom models fit specialized layouts.

Why does Azure OpenAI return 429 rate limit errors?▼

429 errors occur when requests exceed your deployment's tokens-per-minute quota. Implement exponential backoff retries, reuse client instances, and monitor token usage via the completion Usage property to manage throughput.