skillgrab

The tool scans a project’s configuration files—such as package.json, requirements.txt, pubspec.yaml, Gemfile, go.mod, Dockerfile, and README—to infer the technologies and non‑code needs (e.g., marketing, SEO, design) being used. Based on the detected signals it queries the skills.sh registry for matching AI skill packages and proposes an installation plan, optionally previewed with a dry‑run flag.

It then runs `npx skills add` for each selected skill, installing them into any detected AI coding agents (e.g., Claude Code, Cursor, Cline, Codex, Gemini, GitHub Copilot, etc.) or a specific agent supplied via `--agent`. The process works across a wide range of front‑end, back‑end, mobile, infrastructure, and auxiliary domains, covering frameworks like React, Next.js, Vue, Svelte, Django, FastAPI, Go, Rust, Ruby on Rails, and services such as Supabase, Firebase, Stripe, Docker, Vercel, and Cloudflare.

The command‑line interface is intended to be a single step—run inside the project directory, let it scan, suggest, and install the appropriate AI skills, and keep them up to date with subsequent subcommands. It is positioned for developers who want to automate the discovery and integration of AI‑enhanced tooling without manual configuration.