Back to search
skills-management
SkillGitHub
A CLI tool to manage and synchronize AI coding agent skills
Safety scan
A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.
Checked today
Badge for your README
Markdown
[](/listings/nnnggel--skills-management)HTML
<a href="/listings/nnnggel--skills-management"><img src="/api/listings/nnnggel--skills-management/badge.svg" alt="findsafeskills safety rating" /></a>Install
i
Install instructions from the README
As written by the repo's author. findsafeskills has not verified these steps.
This tool can be installed directly from npm:
npm install -g skills-management
Or install from source:
# Clone the repository git clone https://github.com/nnnggel/skills-management.git cd skills-management npm install npm run build npm link
What was scanned
The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.
README.md · README · 156 lines
View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)
- # Skills Management CLI (skm)
- [简体中文](./README.zh-CN.md) | English
- `skm` is a powerful CLI tool designed to manage and synchronize "skills" (prompt libraries, instruction sets, or capability modules) across various AI coding agents and projects. It serves as a central hub to add skills from GitHub or local directories and selectively link them into your local AI project configurations.
- It supports multiple AI environments including **OpenCode**, **Cursor**, **Gemini**, **Antigravity**, **Claude**, and **GitHub** projects.
- ---
- ## 🚀 Features
- - **Global Skill Repository**: Centralized management of your AI skills.
- - **GitHub Integration**: Add skills directly from GitHub repositories or specific subdirectories (sparse checkout).
- - **Local Skill Support**: Add skills from local directories for development and debugging.
- - **Version Control**: Check for updates and synchronize changes from remote repositories.
- - **Project Detection**: Automatically detects the AI project type in your current directory.
- - **Symbolic Linking**: Efficiently links skills to projects without duplication, keeping them in sync.
- - **Project Isolation**: Manage different sets of skills for different projects.
- 
- ## 💡 Why skm?
- - **Centralized Efficiency**: Skills are cloned only once (`~/.skills-management/repo`) and shared across infinite projects via symbolic links. This saves significant disk space and ensures all your projects use the same curated version of a skill.
- - **Precision (Sparse Checkout)**: Only download what you need. If a GitHub repository contains hundreds of skills but you only want one, `skm` uses Git's sparse-checkout to download only that specific subdirectory. No more cloning massive repositories for a single prompt file.
- - **Automatic Environment Awareness**: You don't need to know where Cursor, Claude, or OpenCode stores their skills. `skm` automatically detects your project environment and handles the directory structures for you.
- - **Clean & Non-Intrusive**: Since skills are symlinked, your project folder stays clean. No extra `.git` folders or heavy files are added to your project's version control.
- - **Self-healing**: Proactively detects and helps you clean up broken symbolic links if a global skill is deleted.
- - **Version Sync**: Check for remote updates to keep your local "AI brain" sharp.
- ## 📦 Installation
- This tool can be installed directly from npm:
- ```bash
- npm install -g skills-management
- ```
- Or install from source:
- ```bash
- # Clone the repository
- git clone https://github.com/nnnggel/skills-management.git
- cd skills-management
- npm install
- npm run build
- npm link
- ```
- ## 📖 Usage
- Run the tool using `skm`:
- ```bash
- skm
- ```
- ### 1. Global Repository Management (`repo`)
- Select **"1. repo"** from the main menu to manage your global collection of skills.
- #### Add Skill
- **GitHub Type**:
- - Full repositories: `https://github.com/user/repo` (example: https://github.com/blader/humanizer)
- - Specific subfolders: `https://github.com/user/repo/tree/main/path/to/skill` (example: https://github.com/anthropics/skills/tree/main/skills/pdf)
- **Local Type**:
- - Local directories: `/local/path/to/skill` (example: `~/Desktop/myskill`)
- #### Other Operations
- - **Update Skills**: Checks for newer commits on the remote GitHub repository and updates your local copy (GitHub type only).
- - **Delete Skills**: Remove skills from your global repository.
- ### 2. Project Skill Management (`list`)
- Navigate to your AI project directory and run `skm`. The tool will detect the project type (e.g., `.opencode`, `.cursor`).
- Select **"2. list(type)"** to manage skills for the current project.
- - **Link/Unlink**: You will see a list of available global skills.
- - **Checkbox Interface**: Use `Space` to select or deselect skills, and `Enter` to confirm.
- - **Symlinks**: Selected skills are symlinked contents from your global repo into your project's skill directory (e.g., `.opencode/skills/`).
- ### Supported Project Structures
- `skm` automatically detects and installs skills into these directories:
- | AI Type | Skills Installation Path |
- |---------|---------------------------|
- | **Antigravity** | `.gemini/antigravity/global_skills/skills` or `.agent/skills` |
- | **GitHub** | `.copilot/skills` or `.github/skills` |
- | **Cursor** | `.cursor/skills` |
- | **Claude** | `.claude/skills` |
- | **OpenCode** | `.opencode/skills` |
- ---
- ## 🏗️ Directory Structure & Internals
- `skm` stores its global data and configuration in your home directory at `~/.skills-management`. This architecture ensures that skills are downloaded once and shared across multiple projects.
- ### Data Layout:
- - **`config.json`**: Global configuration file. It stores system-level settings and metadata.
- - **`repo/`**: The heart of the management system.
- - **`skills.json`**: The skill registry. It tracks all added skills, their unique IDs, current commit hashes (for version control), and original paths.
- - **`[type]__[name]/`**: Local skill repository storage. `skm` flattens the ID (replacing `/` with `__`) to create safe directory names. For example, `github:user/repo/path` becomes `github__user__repo__path`, and `local:myskill` becomes `local__myskill`.
- - **Note**: `skm` handles complex paths like `github:user/repo/path/to/skill` by creating a unique folder like `github__user__repo__path__to__skill`.
- ### Cross-Platform Compatibility (Windows)
- `skm` is fully compatible with Windows.
- - On **macOS/Linux**, it uses standard symbolic links.
- - On **Windows**, it automatically uses **Directory Junctions** (a type of symbolic link for folders) to ensure capability without requiring Administrator privileges or Developer Mode.
- ### Custom AI Tools Configuration
- You can customize the AI tools detection by adding `aiTools` to your `~/.skills-management/config.json`:
- ```json
- {
- "system": "darwin",
- "aiTools": [
- { "type": "cursor", "skillDirs": [".cursor/skills"] },
- { "type": "custom-ai", "skillDirs": [".custom-ai/skills", ".custom-ai/prompts"] }
- ]
- }
- ```
- Each entry requires:
- - **`type`**: A unique identifier for the AI tool (used in menu display)
- - **`skillDirs`**: An array of possible skill directory paths (relative to project root)
- > **Note**: When `aiTools` is configured, it completely replaces the default configuration. If you want to keep some defaults, include them in your custom configuration.
- ---
- ## 📝 Changelog
- ### v1.0.5
- - Display version number in main menu
- - Show non-skm managed skills when managing project skills
- - Support custom AI tools configuration via `config.json`
- ### v1.0.4
- - Adjusted AI tool detection directories to match official configurations
- ### v1.0.3
- - Local Skill Support: Add skills directly from local directories
- - Menu Navigation Optimization: Unified numeric menu style, added return options, improved terminal compatibility
- ## 📄 License
- ISC
Ratings and comments
Sign in to rate this skill.
Sign inFree to use. If it’s helped, pass it on.