Agentic Resource Discovery

Assess

Platforms

A draft open specification for discovering and searching agent tools, skills, and other agents across federated registries.

Why it's here

Placed in Assess: 3 article(s) of evidence from 3 source(s), led by framework updates, with 0 in the last 30 days. Confidence 54%.

7The New Stack·8/31/2026framework_update
ARD aims to add discovery to MCP-style agent tooling
Agentic Resource Discovery (ARD) is an open specification intended to help AI agents find tools and resources across multiple registries, clouds, and platforms instead of requiring preconfigured endpoints. The proposal, authored by contributors from Google, Microsoft, and Hugging Face and discussed by AWS, uses JSON-LD and REST to enable federated search for "agentic resources."
7InfoQ·7/14/2026framework_update
Google and Partners Unveil Agentic Resource Discovery Spec
Google and industry partners announced Agentic Resource Discovery (ARD), an open standard for publishing, discovering, and verifying AI tools, APIs, and agents. The specification adds a discovery layer based on catalogs and registries, while relying on existing protocols like MCP and OpenAPI for execution to improve trust and interoperability.
6Hugging Face Blog·6/17/2026framework_update
Hugging Face Introduces Agentic Resource Discovery
Hugging Face describes Agentic Resource Discovery (ARD), a draft open specification for cataloging and searching tools, skills, and agents across federated registries. It positions ARD as a discovery layer in front of MCP, Skills, and A2A, and says its Hugging Face Discover Tool is a reference implementation for searching Spaces, MCP servers, and agent skills at runtime.