One ruleset to sync across 20+ AI coding tools
Rulesync generates tool-specific configs from unified AI rule files, so you don't have to rewrite the same instructions for Cursor, Claude Code, Copilot, and two dozen other agents.

What it does
Rulesync is a Node.js CLI that acts as glue between the growing pile of AI coding tools. You maintain unified rule files—covering behavior, MCP servers, skills, subagents, and ignore patterns—and the tool generates the correct, tool-specific configuration files for each target. It can also import existing configs (like .cursorrules or CLAUDE.md) into its format, or convert directly between two tools without ever creating a .rulesync/ directory.
The interesting bit
The real work is in the mapping. The README’s compatibility matrix spans more than 20 tools, tracking not just simple text rules but nuanced features like hooks, permissions, and whether a capability works in project mode, global mode, or only as a simulated equivalent. It even keeps deprecated targets alive—Google’s Gemini CLI, slated for retirement in June 2026, remains supported because enterprise repositories still rely on it.
Key highlights
- Covers 20+ tools including Cursor, Claude Code, GitHub Copilot, Codex CLI, Zed, Windsurf, and Cline.
- Handles rules, commands, MCP, ignore files, subagents, skills, hooks, and permissions where the target supports them.
- The
importsubcommand ingests existing tool configs; theconvertsubcommand migrates between tools in one shot. - Tracks project vs. global vs. simulated modes per feature, so generated configs match each tool’s expectations.
- Maintains deprecated aliases (Gemini CLI, legacy Antigravity) for backward compatibility.
Caveats
- Feature coverage is uneven; the compatibility matrix shows many blank cells, and some capabilities are restricted to global or simulated modes.
- Two Google targets (Gemini CLI and legacy Antigravity) are deprecated, though Rulesync retains them because existing repositories and enterprise plans still depend on them.
Verdict
A solid utility for teams standardizing rules across multiple AI agents or migrating between editors. If you’re monogamous with one tool, you likely don’t need the extra abstraction layer.
Frequently asked
- What is dyoshikawa/rulesync?
- Rulesync generates tool-specific configs from unified AI rule files, so you don't have to rewrite the same instructions for Cursor, Claude Code, Copilot, and two dozen other agents.
- Is rulesync open source?
- Yes — dyoshikawa/rulesync is open source, released under the MIT license.
- What language is rulesync written in?
- dyoshikawa/rulesync is primarily written in TypeScript.
- How popular is rulesync?
- dyoshikawa/rulesync has 1.4k stars on GitHub and is currently cooling off.
- Where can I find rulesync?
- dyoshikawa/rulesync is on GitHub at https://github.com/dyoshikawa/rulesync.