Custom Skills

AI Agent Skills

This repository contains custom skills for AI agents, providing specialized capabilities and integrations for Swift and the Apple ecosystem.

Available Skills

SwiftZilla

Search Swift technical context using SwiftZilla Deep Insight™ and analyze project dependencies with depgraph.

Location: skills/swiftzilla/

Swift Best Practices & Ecosystem

We also offer curated knowledge base skills for the Swift ecosystem:


🚀 How to Install Skills

These skills are compatible with the Agent Skills standard (like the open-source Vercel skills.sh ecosystem) and can be used on AI assistants like Cursor, Cline, or Context7.

Option 1: Using skills.sh (npx)

If your agent environment supports skills.sh, you can add these skills directly using the Vercel skills CLI format. For example, to install the swiftzilla skill:

npx skills add SwiftZilla/skills

*(Adjust the owner/repo path to match the GitHub URL where this repository is hosted).*

Option 2: Using Context7

If you use Context7 or similar platforms that support importing agent skills via repositories, you can directly import this GitHub repository or point the platform configuration to the specific skill directory you want to use (e.g., SwiftZilla/skills).

Option 3: Manual Installation

To install a skill manually into an AI agent workspace (e.g., Cursor, Deepmind/Cline environments, etc.):

1

Create a .agents/skills or _agents/skills directory inside your project root.

2

Manually copy the skill directory (e.g., skills/swiftzilla) and its contents into your project's skills folder.

3

Make sure to read the SKILL.md inside each skill for any additional setup (e.g., setting up the SWIFTZILLA_API_KEY, granting execution permissions with chmod +x for binaries, etc.).


*Add new skills by creating a subdirectory in skills/ with a SKILL.md file following the skill specification format.*