My Git is a handbook and educational resource that teaches Git principles and workflows for AI-native software engineering.
The handbook addresses the gap between AI tools automating Git commands and developers understanding the concepts behind those commands. It argues that while AI can execute operations, humans need to grasp Git's underlying design to make informed decisions about what changes to accept, what questions to ask, and how to verify results. The approach shifts learning focus from command memorization toward understanding design principles: how snapshots record state, how objects reuse content, how references organize history, and how teams handle conflicting changes. The resource provides interactive learning tools, structured learning paths, and practical guidance organized around three levels of need: understanding Git's design, completing individual engineering changes, and managing team collaboration.
The handbook suits Git newcomers building mental models, developers frequently handling conflicts and rollbacks, advanced developers managing pull requests and CI workflows, technical leads designing team Git practices, and AI coding tool users who need to control and verify code changes from tools like Codex, Claude, Cursor, and Cline. It explicitly covers both what Git can do and its boundaries, noting that complete change control requires Git alongside testing, code review, permissions, artifact management, and release systems. The resource includes a knowledge map to help readers identify whether they need foundational design understanding, guidance on single changes, or team collaboration patterns.
The project maintains structured documentation with interactive components, including a mental model visualizer and organized learning paths. The repository includes practical examples such as AI change review samples and GitHub engineering governance guidelines. Content is available in both English and Chinese, reflecting attention to accessibility across language communities.
