AWS launches generally available Agent Registry for runtime tool discovery
AWS announced that its Agent Registry service is generally available, offering a managed, searchable catalog of Model Context Protocol (MCP) servers, agents, and agent skills. Publishers register protocol‑specific descriptors and searchable descriptions, while orchestrators query the registry at runtime to discover the appropriate capability, instantiate a client, and execute the request.
Key points
- AWS Agent Registry provides a governed, searchable catalog for MCP servers, agents, and skills, enabling runtime discovery.
- The sample demonstrates a discover‑instantiate‑execute loop using Bedrock AgentCore, Cognito, Lambda, and other AWS services.
- Runtime discovery decouples tool publishing from orchestrator code, allowing new capabilities without redeploying the orchestrator.
The blog post walks through a Python sample that deploys an order‑management MCP server through Amazon Bedrock AgentCore Gateway, two agent‑to‑agent (A2A) agents, the Registry, and an orchestrator. Three demo scenarios illustrate the discover‑instantiate‑execute loop: an MCP‑only order status query, a combined MCP + A2A pricing and discount flow, and a semantic match for a return‑and‑refund intent that does not appear in any tool name. The sample shows how authentication (Cognito OAuth for Gateway, IAM‑signed Boto3 calls for A2A) remains separate from discovery, and how governance via approval workflows controls which records become discoverable.
The story so far
5 episodes →- AWS launches generally available Agent Registry for runtime tool discoverythis story
Discover tools at runtime with AWS Agent Registry
AWS Machine Learning Blog · 20 October 2026
Loading the full article…
This text was published by AWS Machine Learning Blog and written by Vibhu Pareek. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
4sources- Build a semantic layer for agentic AI on AWS with Denodo and Amazon Bedrock AgentCorePrimary source · AWS Machine Learning Blog ·
- Build a semantic layer for agentic AI on AWS with Strategy Mosaic and Amazon Bedrock AgentCorePrimary source · AWS Machine Learning Blog ·
- Build a semantic layer for agentic AI on AWS with AtScale and Amazon Bedrock AgentCorePrimary source · AWS Machine Learning Blog ·
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Agents & Tools
All →- Anthropic’s MCP protocol lets AI use Slack, Google Drive and browsers via plain-language commands · 1 src
- AWS and Keyfactor detail X.509 PKI architecture for securing AI agents · 1 src
- Eon uses Amazon Bedrock AgentCore gateway to unify multi-cloud investigation tools · 1 src
- Amazon releases TorchNeuron to run PyTorch on Trainium chips, challenging Nvidia's CUDA advantage · 5 src
- Cloudflare and OpenAI pilot indexes hundreds of millions of new pages for ChatGPT · 1 src
Comments
via GitHub Discussions