This dataset contains data on repository-level configuration artifacts for five agentic AI coding tools (Claude Code, Gemini CLI, Codex CLI, Copilot CLI, Cursor) across 4,738 GitHub repositories.





The full pipeline used to create this dataset is available on Zenodo:


Sebastian Baltes, Seyedmoein Mohsenimofidi, Levi Böhme, Jai Lal Lulla, Muhammad Auwal Abubakar, Christoph Treude, Matthias Galster. A Dataset of Agentic AI Coding Tool Configurations (Pipeline). Zenodo, Apr. 01, 2026. doi: 10.5281/zenodo.19392306.This dataset was accepted in the AIWare 2026 Datasets and Benchmarks track.Claude Code (Opus 4.6, high reasoning effort) was used to write and revise code and documentation for the pipeline that produced this dataset. The research design and methodology are the work of the human authors.