Overview

The skills CLI is your gateway to the open agent skills ecosystem, providing a seamless way to discover, install, and manage reusable instruction sets for your coding agents. It empowers your AI assistants with specialized capabilities, from generating release notes to integrating with your favorite tools, all without leaving your terminal.

Key Features

• Effortless Skill Management: Install, list, find, update, and remove skills for a wide range of supported agents (including OpenCode, Claude Code, Codex, Cursor, and many more) with simple, intuitive commands.
• Flexible Source Options: Install skills from GitHub (shorthand or full URL), GitLab, local paths, or even directly from download URLs. Supports private repositories with your existing Git authentication.
• Agent-Specific Installation: Target specific agents or install globally, choosing between project-level integration or system-wide availability.
• Interoperability: Use skills without installation by piping their output to supported agents.
• Custom Skill Creation: Easily define your own agent skills using the SKILL.md format, complete with frontmatter for name, description, and custom instructions.
• Extensive Agent Support: Integrates with a vast and growing list of AI coding agents, ensuring your skills can be leveraged wherever you work.
• Intelligent Discovery: Automatically detects installed agents and intelligently discovers skills within repositories based on standard structures.

Typical Use Cases

• Enhance Coding Assistants: Grant your AI pair-programmer specialized knowledge for tasks like code review, test generation, or API integration.
• Automate Repetitive Tasks: Streamline workflows by creating skills for tasks such as generating release notes from Git history or formatting code according to specific standards.
• Integrate with External Tools: Build bridges between your coding agent and services like Linear, Notion, or other project management and documentation platforms.
• Share Team Knowledge: Develop and distribute custom skills within your team to enforce best practices and standardize common development patterns.
• Experiment with New Agent Capabilities: Quickly test and integrate new AI functionalities by discovering and applying community-contributed skills.