rulesync

Configure and generate AI assistant rules, permissions, and MCP settings from unified sources.

1|Updated Sep 3, 2017
One-click install
npx skills add https://github.com/mmgeorge/config --skill rulesync-mmgeorge
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: rulesync
Source: https://github.com/mmgeorge/config/tree/main/rulesync-global/.rulesync/skills/rulesync
Command: npx skills add https://github.com/mmgeorge/config --skill rulesync-mmgeorge

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Managing configuration files for multiple AI coding assistants (Claude Code, Cursor, Copilot, Codex CLI, and others) means duplicating rules, permissions, and MCP server settings across incompatible formats. This Skill helps you edit Rulesync JSONC configuration correctly, select the right target-feature combinations, and diagnose why generated files are missing or stale. ## Core Features & Use Cases - Configuration Management: Edit rulesync.jsonc and rulesync.local.jsonc to control which providers and features (rules, permissions, MCP, skills, commands, hooks) are generated, including object-form per-target feature selection. - Generation and Import Workflows: Run rulesync generate, rulesync import, rulesync fetch, and rulesync install with correct input roots, global scope, and dry-run validation. - Diagnosing Stale Output: Trace missing generated files back to feature selection, source discovery, or scope mismatches instead of editing generated files directly. - Use Case: You changed .rulesync/permissions.json but Codex CLI permissions did not update. The Skill guides you to check whether targets.codexcli includes the permissions feature before regenerating. ## Quick Start Ask the assistant to review your rulesync.jsonc configuration and run a dry-run generation to verify the intended target-feature output.

Frequently Asked Questions about rulesync

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate AI assistant configuration files with Rulesync?▼

Run rulesync generate after defining targets and features in rulesync.jsonc. Use rulesync generate --dry-run first to preview changes, then generate to write files like CLAUDE.md, .cursorrules, or .github/copilot-instructions.md.

How do I enable a specific feature for one AI tool in rulesync.jsonc?▼

Use the object form of targets, for example "targets": { "codexcli": ["rules", "permissions"] }. When targets is an object, the top-level features field must be omitted because the config loader rejects defining both.

Why is my generated file missing after editing a .rulesync source?▼

Editing a source asset alone does not enable its feature for a provider. Check that the provider's target entry includes that feature, verify the selected config file and input root, then run a dry run and inspect skipped-feature warnings.

Does Rulesync support global user-scope configuration?▼

Yes, enable global mode with the --global flag or "global": true in rulesync.jsonc. Global mode currently supports rules generation for Claude Code, GitHub Copilot, and OpenCode, with command generation limited to Claude Code.

What is the difference between rulesync generate --dry-run and --check?▼

Both preview changes without writing files. The --check flag additionally exits with code 1 when generated files are not up to date, making it suitable for CI/CD pipeline validation.

How do I install skills from external repositories with Rulesync?▼

Declare a sources array in rulesync.jsonc and run rulesync install, which resolves refs into a rulesync.lock lockfile and fetches skills into .rulesync/skills/.curated/. Local skills in .rulesync/skills/ take precedence over curated ones.