jMRI (Just My Retrieval Interface) is a minimal specification for efficient code retrieval in AI agents, designed to replace chunk-based approaches like those used in Chunked Code Compression (CCC). It includes a Python and TypeScript SDK, and its core interface focuses on discovering indexed sources, searching for specific symbols, and retrieving exact symbol definitions. The jMRI v1.0 spec is available under Apache 2.0 license at github.com/jgravelle/mcp-retrieval-spec. Implementing jMRI-Core in retrieval MCP servers can significantly reduce token usage and improve performance for AI agents like Claude Code, Cursor, and generic agents.
Read the full article at DEV Community
Want to create content about this topic? Use Nemati AI tools to generate articles, social posts, and more.
