Aggregates multiple Model Context Protocol (MCP) servers into a single local endpoint for AI assistants.
Sponsored
Endara Relay is a local Rust binary designed to simplify the interaction between AI assistants (like Claude Desktop or Cursor) and various Model Context Protocol (MCP) servers. It acts as a central hub, allowing users to point their AI client to a single local endpoint, which then seamlessly connects to and manages multiple MCP servers, regardless of their underlying transport (STDIO, SSE, or HTTP). This tool merges diverse tool catalogs into a unified list, automatically prefixes tool names to prevent conflicts, and ensures high availability through features like hot-reloading configuration changes and automatic crash recovery for server processes. Everything runs securely on the user's machine, eliminating the need for cloud services or accounts.
주요 기능
01Unifies and prefixes tool catalogs to prevent naming collisions
02Automatically restarts crashed STDIO server processes with exponential backoff
03Offers a JavaScript execution mode to reduce AI context window pollution with meta-tools
04Aggregates multiple MCP servers (STDIO, SSE, HTTP) behind a single endpoint
050 GitHub stars
06Hot-reloads configuration changes without requiring a restart
사용 사례
01Simplifying the setup and management of multiple AI tool backend services
02Consolidating diverse local and remote MCP servers for a unified AI assistant experience
03Enabling AI assistants to dynamically discover and execute tools through a secure, in-process JavaScript sandbox
