explaingit

vercel-labs/skills

📈 Trending★ 19,109TypeScriptAudience · developerComplexity · 2/5ActiveSetup · easy

TLDR

A command-line tool for installing and managing reusable instruction sets (skills) that teach AI coding agents new capabilities and guidelines.

Mindmap

mindmap
  root((repo))
    What it does
      Install skills
      Manage versions
      Share guidelines
    How it works
      Markdown files
      Agent directories
      Symlinks or copies
    Installation sources
      GitHub repos
      GitLab repos
      Local directories
      Git URLs
    Use cases
      Team coding standards
      Agent customization
      Shared instructions
    Scope options
      Project level
      Global machine
    Key commands
      Add skills
      List installed
      Search skills
      Update versions
      Remove skills
      Create template

Things people build with this

USE CASE 1

Install coding standards and best practices as reusable skills that all team members' AI agents follow automatically.

USE CASE 2

Create custom instruction sets for your AI agent to handle project-specific patterns, naming conventions, or architectural rules.

USE CASE 3

Share and discover pre-built skills from the community to extend your agent's capabilities without writing code.

USE CASE 4

Manage multiple versions of skills across projects and keep them synchronized with symlinks or file copies.

Tech stack

TypeScriptNode.jsnpm/npx

Getting it running

Difficulty · easy Time to first run · 5min
License could not be detected automatically. Check the repository's LICENSE file before use.

In plain English

This repository contains a command-line tool (run via npx) for managing "agent skills", reusable instruction sets that extend the behavior of AI coding agents. A skill is defined as a markdown file with some metadata at the top, and once installed, the coding agent reads these files to gain new capabilities or follow specific guidelines. The tool supports a wide range of coding agents and works by installing skill files into the appropriate directory each agent looks in for instructions. Skills can be installed from a GitHub or GitLab repository, a local directory, or a direct git URL. They can be installed for a single project (committed alongside project code so a whole team shares them) or globally for all projects on a machine. The main commands let users add skills from a source, list what is installed, search for available skills, update installed skills to newer versions, remove skills, and create a template for writing a new skill. Skills can be installed as symlinks (so all agents share one copy that is easy to update) or as file copies (for environments where symlinks are not supported).

Copy-paste prompts

Prompt 1
How do I install a skill from a GitHub repository using this tool, and where does it get stored on my machine?
Prompt 2
Show me how to create a new skill template and what metadata I need to include in the markdown file.
Prompt 3
I want my whole team to use the same coding guidelines in their AI agents, how do I install skills at the project level so they're committed to git?
Prompt 4
What's the difference between installing skills as symlinks versus file copies, and when should I use each?
Prompt 5
How do I search for available skills and update an installed skill to a newer version?
Open on GitHub → Explain another repo

Generated 2026-05-18 · Model: sonnet-4-6 · Verify against the repo before relying on details.