Discover tools at runtime with AWS Agent Registry
Hardcoded tool wiring breaks down as an agent's catalog grows from a handful of tools to hundreds. This post walks through a runtime Discover, Instantiate, Execute loop built on AWS Agent Registry, where an orchestrator discovers MCP servers and A2A agents through semantic search at runtime instead of compiling every tool at build time.
AWS Agent Registry is a managed service that allows enterprise software SaaS providers to catalog and discover Model Context Protocol (MCP) servers, agents, and agent skills. The service enables runtime discovery and instantiation of tools for agentic consumers. Instead of hardcoding tool names and endpoints, the AWS Agent Registry enables orchestrators to search for and retrieve discoverable records at runtime.
The orchestrator can then use the returned endpoint and descriptor content to construct a client for the requested tool or agent. This approach eliminates the need for code changes and redeployment when new capabilities are added, and allows teams to ship updates on their own cadence. The AWS Agent Registry provides a governed, searchable catalog of MCP servers and agents, with each record containing searchable descriptions and protocol-specific descriptors.
The records do not include invocation credentials, which must be obtained separately through a separate, trusted configuration path. The registry discovery process does not return authentication material, ensuring secure access to the tools and agents. The AWS Agent Registry can be integrated with other AWS services, such as Amazon Bedrock AgentCore Gateway and Amazon Bedrock AgentCore runtime, to enable end-to-end orchestration of internal APIs and capabilities.
Written by urgent.news from AWS Machine Learning's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
