Toolboxes in Foundry: Build, Search and Govern Your Tools | LIVE163

Seth Juarez explains how Azure AI Foundry Toolboxes let teams build, discover, and govern tools across multiple AI agents, reducing duplicated integration work around authentication, credentials, and endpoint wiring.

Overview

Toolboxes in Azure AI Foundry are presented as a way to avoid each AI agent wiring tools independently (including separate authentication, credentials handling, and integration code). Instead, teams can centralize tool definitions and make them discoverable and reusable across agents.

Key ideas highlighted in the session:

• Centralized tool building and reuse

  • Build tools once and make them available to multiple agents.
  • Reduce repeated integration work when agents need access to the same capabilities.

• Runtime tool search

  • Agents can discover tools at runtime via tool search, rather than being hard-wired to a fixed set of tools.

• Unified endpoint approach

  • Toolboxes provide a unified endpoint model so agents can access tools consistently.

• Governance and guardrails

  • Toolboxes are positioned as a governance layer for tools used by agents.
  • Guardrail integration is called out as part of the toolbox approach.

• Authentication and credentials concerns

  • The session frames authentication/credentials as a core pain point when each agent integrates tools separately.

Speakers

• Atul Aggarwal
• Seth Juarez
• Zhuoqun Li

Resources

• https://aka.ms/Foundry/Toolbox/Docs
• https://aka.ms/toolbox-build-blog
• https://aka.ms/Foundry/AgentService

Session context

This session is part of Microsoft Build 2026 (LIVE163).