Your AI skills are scattered; this Mac app herds them
It exists to stop you digging through dotfiles to manage coding agent skills across half a dozen AI tools.

What it does
Chops is a native macOS app that scans your home directory for AI skills and agents scattered across Claude Code, Cursor, Codex, Windsurf, Amp, and others. It presents them in a three-column browser with a built-in monospaced editor, so you can read, search, and tweak prompts without touching dotfiles directly. It also watches the filesystem in real time, updating its index automatically when something changes on disk.
The interesting bit
The app intentionally disables the macOS sandbox so it can freely roam ~/ and read every tool’s hidden configuration directories—a rare, deliberate choice for a modern Mac app. It deduplicates skills by resolving symlinks, so the same file linked into multiple tool directories appears once with multiple tool badges rather than as redundant copies.
Key highlights
- Supports Claude Code, Cursor, Codex, Windsurf, Amp, Copilot, and Aider, scanning their dotfile or project paths for skills and agents
- Real-time FSEvents-based file watching and in-memory full-text search across names, descriptions, and content
- Collections let you organize skills without moving source files, and it generates correct boilerplate for new skills per tool
- Connects to remote servers running OpenClaw or Hermes to browse and install skills
- Native SwiftUI + SwiftData with zero web views; wraps
NSTextViewfor editing
Caveats
- No automated test suite; the README explicitly tells contributors to validate changes manually
- Disables the macOS app sandbox entirely, which is necessary for core functionality but a notable security tradeoff
- Requires macOS 15 (Sequoia) or later, limiting it to relatively recent hardware
Verdict
Grab it if you juggle multiple AI coding tools and are tired of context-switching between hidden config directories. If you only use one agent or prefer editing dotfiles in Vim, it solves a problem you don’t have.
Frequently asked
- What is Shpigford/chops?
- It exists to stop you digging through dotfiles to manage coding agent skills across half a dozen AI tools.
- Is chops open source?
- Yes — Shpigford/chops is an open-source project tracked on heatdrop.
- What language is chops written in?
- Shpigford/chops is primarily written in Swift.
- How popular is chops?
- Shpigford/chops has 1.9k stars on GitHub and is currently accelerating.
- Where can I find chops?
- Shpigford/chops is on GitHub at https://github.com/Shpigford/chops.