Show HN: shadowclone, skills learned from your coding agent corrections (github.com)

shadowclone has launched as a tool designed to standardize coding preferences across multiple AI coding agents, including Claude Code, Codex, Cursor, Pi, and Antigravity. By enabling users to maintain a consistent set of skills and instructions, shadowclone streamlines the coding process without duplicating rules between platforms. Its ability to learn from user sessions allows it to refine and adapt preferences over time, making it a powerful addition for developers who work across various coding tools. The installation and configuration process is user-friendly and quick, taking approximately five minutes. This innovation is significant for the AI/ML community as it enhances collaboration and efficiency in coding environments. By integrating a GitHub bot that replicates user-selected skills, shadowclone can handle code writing and reviews that reflect a user's personalized guidance both locally and in the cloud. Furthermore, its built-in learning mechanism ensures that guidance remains relevant and up-to-date, thus improving adherence to coding standards. The tool's architecture focuses on privacy, as it learns without saving session transcripts and allows for easy edits and reversibility of rules. Overall, shadowclone aims to boost productivity for developers by providing a cohesive skill set across multiple AI platforms.