MCP-Grep provides a server implementation that exposes the system grep binary's functionality through the Model Context Protocol, allowing AI assistants to search for patterns in files using regular expressions. Developed by Ernie Brodeur, it offers both a resource endpoint for retrieving grep binary information and a tool for performing searches with support for common grep options like case-insensitive matching, context lines, recursive directory searching, and fixed string matching. The implementation uses Python's subprocess module to execute grep commands on the host system, making it particularly valuable for AI assistants that need to analyze text files or codebases without leaving their conversation interface.
erniebrodeur