dyoshikawa/rulesync

What it solves

Rulesync is a CLI tool designed to eliminate the redundancy of maintaining separate configuration files for multiple AI coding assistants. Instead of manually updating rules, commands, and MCP settings across different tools (like Cursor, Claude Code, and GitHub Copilot), developers can maintain a single source of truth in unified rule files and automatically generate the specific configurations required by each tool.

How it works

Rulesync uses a .rulesync/ directory as the central repository for AI rules. It provides a set of commands to manage these rules:

• Generation: The generate command converts unified rules into the native configuration formats of various supported AI tools.
• Importing: The import command allows users to migrate existing tool-specific configurations (e.g., .cursorrules or CLAUDE.md) into the unified Rulesync format.
• Conversion: The convert command can directly transform configurations from one AI tool to another without needing to adopt the .rulesync/ workflow.
• Fetching: The fetch command allows users to install official skills or pre-defined rulesets.

Who it’s for

It is primarily for developers who use multiple AI coding tools simultaneously or switch between them, and those who want a consistent AI behavior across their entire development environment.

Highlights

• Broad Tool Support: Supports a massive array of AI coding tools including Claude Code, Cursor, GitHub Copilot, and many others.
• Unified Configuration: Manages rules, ignore files, MCP (Model Context Protocol) settings, commands, subagents, and skills in one place.
• Comprehensive Migration: Offers both import and direct conversion capabilities to move between tool ecosystems.
• Flexible Generation: Supports selective generation based on targets and features, with support for both project-level and global configurations.