MMCPMarkets
AS

auran0s/sklm

Developer tools

このページは AI により自動翻訳されています。技術的な詳細は原文をご確認ください。 原文を見る

11 stars 品質 68 トレンド 26MIT
Repository Install

Sklm is a CLI tool that centralizes management of skills (SKILL.md files) for your AI agents. It solves the tension between wanting skills globally available vs.

概要

Sklm lets you install AI agent skills once and activate them per project. No more copy-pasting SKILL.md files between directories. - and every project sees it, even when it's irrelevant. - and it's invisible to others. Sklm keeps a global library in ~/.sklm/, then lets you pick which skills each project sees. Sklm only supports skill resources for now. More resource types may come later. - — from OpenCode and Claude Code to Codex CLI, GitHub Copilot, and beyond. - — a global store at ~/.sklm/ holds your skills; per-project symlinks activate only what you need. - — sklm add and sklm rm automatically update the agent's skills directory. No manual copying. - — index local folders or git repos as searchable skill catalogs. - — sklm add --from clones a repo and figures out where the skill lives. - — a single skill can ship agent-specific file overrides in a variants/ subdirectory. Each agent receives the version tuned for it. For development, clone the repo and use pip install -e .

README

以下は原文 README の抜粋です。原文を見る。

Why Sklm?

Sklm lets you install AI agent skills once and activate them per project. No more copy-pasting SKILL.md files between directories.

  • Install a skill globally and every project sees it, even when it’s irrelevant.
  • Drop it in one project and it’s invisible to others.

Sklm keeps a global library in ~/.sklm/, then lets you pick which skills each project sees.

[!NOTE] Sklm only supports skill resources for now. More resource types may come later.

Features

  • Works with 30 AI agents — from OpenCode and Claude Code to Codex CLI, GitHub Copilot, and beyond.
  • Install once, scope per project — a global store at ~/.sklm/ holds your skills; per-project symlinks activate only what you need.
  • Auto-sync — sklm add and sklm rm automatically update the agent’s skills directory. No manual copying.
  • Registry discovery — index local folders or git repos as searchable skill catalogs.
  • Git repo installation — sklm add --from clones a repo and figures out where the skill lives.
  • Per-agent skill variants — a single skill can ship agent-specific file overrides in a variants/ subdirectory. Each agent receives the version tuned for it.

Installation

pip install sklm-cli

[!TIP] For development, clone the repo and use pip install -e . for an editable install.

Quickstart

pip install sklm-cli                   # install globally
sklm                               # interactive wizard opens — detects your setup

That’s it. The CLI’s interactive wizard detects your AI agents, initializes the workspace, and guides you through adding your first skill — no flags needed.

[!TIP] Run sklm init --agent opencode to skip the wizard and set a specific agent. Pass --agent multiple times for multiple agents.

Usage

Workspace setup

sklm init                          # auto-detect or prompt for agent(s)
sklm init --agent opencode         # force a specific agent
sklm init --agent claude --agent cursor   # multiple agents at once
sklm status                        # show workspace health
sklm status --repair               # fix broken symlinks
GitHub で README 全文を見る

推奨ツール

別のキーワードを試すか、フィルタを外してください。

インストール

npx skillfish add auran0s/sklm
MCP 初心者？mcpskill.org のガイドから始めましょう →