skills.rest
microsoft
microsoftOFFICIAL
Explore all skills by this author →
Website𝕏@OpenAtMicrosoftGitHub

deploying-scalable-agents

Configure production-ready agent deployments with observability, evaluation gates, and scalable infrastructure patterns.

73.6k|24.3k|Updated Nov 28, 2024
Search Tags:#Scalability#Deployment#AGENTS#Foundry#openTelemetry#Observability
One-click install
npx skills add https://github.com/microsoft/ai-agents-for-beginners --skill deploying-scalable-agents
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: deploying-scalable-agents
Source: https://github.com/microsoft/ai-agents-for-beginners/tree/main/.agents/skills/deploying-scalable-agents
Command: npx skills add https://github.com/microsoft/ai-agents-for-beginners --skill deploying-scalable-agents

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill addresses the gap between building a functional agent prototype and maintaining a reliable, cost-effective, and observable production service.

Core Features & Use Cases

  • Deployment Patterns: Implement client-hosted, hosted-agent, or agent-workflow architectures using Microsoft Foundry.
  • Operational Controls: Configure model routing, response caching, and human-in-the-loop approval workflows to manage cost and trust.
  • Production Readiness: Integrate OpenTelemetry for tracing and automated evaluation gates to ensure quality before deployment.

Quick Start

Use the deploying-scalable-agents skill to configure an evaluation gate and smoke test for my production agent deployment.

Frequently Asked Questions about deploying-scalable-agents

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

FAQPage Schema
How do I deploy an agent to Microsoft Foundry?▼

Deploy an agent by selecting a deployment pattern such as hosted-agent or agent-workflow, then configuring the agent with managed identity and scoped RBAC within the Foundry environment.

What is an evaluation gate in agent deployment?▼

An evaluation gate is an automated quality check that runs an agent against an offline test set, ensuring it meets a defined pass-rate threshold before it is promoted to production.

How can I optimize agent costs in production?▼

Optimize costs by right-sizing models, implementing model routing to use smaller models for simple tasks, and utilizing response caching to avoid redundant model calls.

Does this skill support local agent development?▼

No, this skill is specifically for production-grade deployments on Microsoft Foundry; for local on-device agent development, refer to the local-ai-agents resources.

How do I monitor agent performance in production?▼

Monitor performance by instrumenting the agent with OpenTelemetry tracing to capture request spans, routed model information, and customer-specific attributes for observability.

Related Skills

Explore More Software Engineering Skills→
tadod12tadod12

chroma

Store embeddings and metadata in a local database for RAG retrieval.

Community
Advanced
abhishekmmgnabhishekmmgn

agent-quality-flywheel

Implement a four-step flywheel converting production data into agent improvements.

Community
Advanced
Probably-GroupProbably-Group
3

mcp

Implement MCP servers with Zod validation, allowlisted commands, and restricted resource access.

Official
Advanced
NovaAI-innovationNovaAI-innovation

devops

Automate cloud infrastructure deployment across Cloudflare, Docker, Google Cloud, and Kubernetes.

Community
Advanced