The "mcp" URI Scheme and MCP Server Discovery Mechanism
draft-serra-mcp-discovery-uri-04
| Document | Type |
Expired Internet-Draft
(individual)
Expired & archived
|
|
|---|---|---|---|
| Author | marco serra | ||
| Last updated | 2026-09-27 (Latest revision 2026-03-26) | ||
| RFC stream | (None) | ||
| Intended RFC status | (None) | ||
| Formats | |||
| Stream | Stream state | (No stream defined) | |
| Consensus boilerplate | Unknown | ||
| RFC Editor Note | (None) | ||
| IESG | IESG state | Expired | |
| Telechat date | (None) | ||
| Responsible AD | (None) | ||
| Send notices to | (None) |
This Internet-Draft is no longer active. A copy of the expired Internet-Draft is available in these formats:
Abstract
The Model Context Protocol (MCP) defines a standard interface for AI agents to connect to external tools and services. However, no standard mechanism exists for an AI agent to autonomously discover which web domains expose an MCP server. This document defines: (1) the "mcp" URI scheme, a machine-to-machine identifier for publicly reachable MCP servers; (2) a two-mode discovery mechanism: a well-known URI (/.well-known/mcp-server) for universal compatibility, and a DNS TXT record format for DNS-native fast discovery; (3) an optional manifest integrity mechanism using JSON Web Signatures (JWS, [RFC7515]) with keys published via DNS; (4) a security capability negotiation mechanism through which the manifest declares trust class, authentication requirements, compliance frameworks, and logging policy before connection. This specification does not define the MCP protocol itself, nor authentication between agent and server. Those are covered by the MCP specification and OAuth 2.1 respectively.
Authors
(Note: The e-mail addresses provided for the authors of this Internet-Draft may no longer be valid.)