New dataset reveals 2,297 real-world MCP projects on GitHub

A new study from researchers Toeppe, Barrak, and Ksontini presents the first large-scale, evidence-based dataset of Model Context Protocol (MCP) implementations collected directly from GitHub. MCP is an emerging standard for connecting large language models to external tools and services, and its rapid adoption in open-source development has outpaced systematic understanding. The team built a hybrid pipeline combining GitHub REST and GraphQL APIs with custom Python verification scripts to discover 3,238 candidate repositories. Through multi-stage evidence checks and manual review of a representative subset, they achieved an overall precision of 83% at a 95% confidence level. After removing non-operational repositories (educational samples, tutorials, demonstrations), the final dataset comprises 2,297 validated MCP projects, exported in a reproducible JSONL schema and classified by operational role (client, server, gateway).

Analysis of the dataset reveals clear development trends: Python and TypeScript dominate MCP implementation, with hybrid architectures emerging as the most common design pattern. By emphasizing transparent verification strategies, structured evidence tagging, and reproducible data organization, this work establishes a foundational benchmark for studying real-world MCP ecosystems. It supports future research on integration, connectivity, and compatibility across the broader developer community, offering a concrete reference point for understanding how developers are adopting and adapting the protocol in practice.