Brand LogoBrand Logo (Dark)
HomeAI AgentsToolkitsGitHub PicksSubmit AgentBlog

Categories

  • Art Generators
  • Audio Generators
  • Automation Tools
  • Chatbots & AI Agents
  • Code Tools
  • Financial Tools

Categories

  • Large Language Models
  • Marketing Tools
  • No-Code & Low-Code
  • Research & Search
  • Video & Animation
  • Video Editing

GitHub Picks

  • DeerFlow — ByteDance Open-Source SuperAgent Harness

Latest Blogs

  • OpenClaw vs Composer 2 Which AI Assistant Delivers More Value
  • Google AI Studio vs Anthropic Console
  • Stitch 2.0 vs Lovable Which AI Design Tool Wins in 2026
  • Monetizing AI for Solopreneurs and Small Teams in 2026
  • OpenClaw vs MiniMax Which AI Assistant Wins in 2026

Latest Blogs

  • OpenClaw vs KiloClaw Is Self-Hosting Still Better
  • OpenClaw vs Kimi Claw
  • GPT-5.4 vs Gemini 3.1 Pro
  • Farewell to Bloomberg Terminal as Perplexity Computer AI Redefines Finance
  • Best Practices for OpenClaw
LinkStartAI© 2026 LinkstartAI. All rights reserved.
Contact UsAbout
  1. Home
  2. GitHub Picks
  3. Skills Desktop
Skills Desktop logo

Skills Desktop

An Electron-based desktop client for visualizing and managing AI agent skills, using symlinks to maintain a unified SKILL.md source across 21 major AI coding assistants.
380TypeScriptUnknown
#electron#ai-skill-manager#symlink-management#multi-agent-support#claude-code-skills
#local-first
#developer-tooling
#alternative-to-openskills
#alternative-to-skills-sh
#cross-agent-workflow

What is it?

Skills Desktop is a native desktop management tool built specifically for developers who heavily rely on AI agent skills. As tools like Claude Code, Cursor, Gemini CLI, OpenAI Codex, and GitHub Copilot all adopt the SKILL.md format, developers increasingly face the same pain point: skill files are scattered across individual tool configuration directories with no unified management. Skills Desktop's core solution is elegant: it centralizes all skills in ~/.agents/skills/ and uses the OS-level symlink mechanism to link each skill into the configuration paths of every supported agent. Edit a skill once locally and every tool is updated instantly, with zero scripts or CI/CD pipelines needed. It also features a real-time visual dashboard that displays the symlink health status (valid, broken, or missing) for every skill across all 21 agents, plus 26 polished OKLCH themes to make skill management an enjoyable daily workflow.

Pain Points vs Innovation

✕Traditional Pain Points✓Innovative Solutions
Managing multiple AI coding assistants means copying the same skill files into each tool's configuration directory manually; any update must be repeated multiple times, causing version inconsistencies.Skills Desktop achieves a true single source of truth by storing all skills in one directory and symlinking to all agents, permanently eliminating the multi-copy sync nightmare.
Existing CLI tools like npx skills add can install individual skills but lack the ability to globally visualize the health status of all local agents, making broken or missing symlinks hard to detect.It provides a native Electron desktop interface and real-time monitoring engine, so any new or modified skill file is instantly reflected on the global status dashboard, catching symlink issues before they cause problems.

Architecture Deep Dive

Symlink-Based Single Source of Truth Architecture
The system's core design philosophy is 'store once, link everywhere.' All SKILL.md files are stored in a unified ~/.agents/skills/ directory as the immutable authoritative source. For each detected AI agent tool, the application creates an OS-level symlink within that tool's designated config directory, pointing back to the central store. The technical advantage of this design is significant: when you edit a file in the central directory, all linked agent paths instantly read the new content. The synchronization is handled entirely by the OS kernel, requiring no background services or network requests, and works flawlessly in offline environments.
21-Agent Auto-Detection and Real-Time Symlink Health Monitoring Engine
The app embeds directory path detection rules for 21 major AI coding assistants, including Claude Code, Cursor, Gemini CLI, OpenAI Codex, and GitHub Copilot. At launch, the engine traverses the local system to auto-discover installed agents and scans the symlink status of each skill across all agents in real-time, classifying them as 'Valid', 'Broken', or 'Missing' on a visual dashboard. Additionally, the system uses a filesystem watcher (FSWatcher) to continuously monitor the ~/.agents/skills/ directory. Any file addition, modification, or deletion triggers an instant dashboard refresh without manual intervention, ensuring the management UI always reflects the actual state of the disk.

Deployment Guide

1. Visit the GitHub Releases page to download the latest Skills Desktop macOS package

bash
1open https://github.com/Harries/skills-desktop/releases

2. Unzip and drag the app to your Applications folder, or install the development build via npm

bash
1npm install -g skills-desktop

3. On first launch, the app auto-scans locally installed AI agent tools and initializes the central skills directory at ~/.agents/skills/

bash
1mkdir -p ~/.agents/skills

4. Use the 'Add Skill' button in the Skills Desktop UI to import or create SKILL.md files; symlinks are automatically created for all detected agents

bash
1echo 'Symlinks established instantly for all detected agents'

Use Cases

Core SceneTarget AudienceSolutionOutcome
Unified Multi-Agent Skill SyncFull-stack developers using multiple AI coding assistantsShare one SKILL.md source across 21 agents via symlinksCompletely eliminate repetitive file copying and sync all agents with a single edit
Skill Asset Health InspectionTech leads maintaining large sets of custom skillsUse the real-time symlink status dashboard to catch broken and missing links across all agentsSurface hidden config failures before they impact development efficiency and fix them quickly
Offline-Stable Local WorkflowIndie developers who travel or work in restricted network environmentsPin all core skills locally without relying on any cloud sync serviceEnsure AI agent skill configurations remain complete and available under any network condition

Limitations & Gotchas

Limitations & Gotchas
  • Currently a native macOS application; Windows and Linux users cannot use it yet—cross-platform support is the most-requested community feature.
  • The entire management capability relies on OS-level symlink support. Running it in filesystems or cloud sync directories (e.g., iCloud Drive) that do not fully support symlinks may cause unexpected behavior.
  • The project is still in active early development; path detection rules for certain AI agents may shift with tool version upgrades, so monitoring GitHub Release Notes for timely updates is recommended.

Frequently Asked Questions

What are the core advantages of Skills Desktop over tools like openskills CLI or skills.sh?▾
openskills CLI and skills.sh are fundamentally command-line installation and distribution tools that solve the problem of 'how to get a skill.' After installation, they do not manage the ongoing status of local skills. Skills Desktop is a local skill asset 'operations and monitoring hub'—it assumes you already have a collection of skills and focuses on 'how to achieve zero-redundancy unified management and real-time health monitoring in multi-agent scenarios.' These are two fundamentally different product layers. They complement each other well: use the CLI to install new skills, and use Skills Desktop to manage and monitor all local skill states.
If one of my AI agents isn't in the 21 supported tools list, can I add it manually?▾
Since the project is in its early stage, whether manual agent path configuration is supported should be confirmed in the latest documentation and release notes. Technically, the symlink mechanism works for any directory—if you know the exact path where your target AI agent reads SKILL.md files, you can manually create the link via terminal commands. That said, it is recommended to submit the path details for your agent tool to the official GitHub Issues to help the community add it to the automatic detection list.
I edited a SKILL.md file but a specific agent isn't picking up the updated content—what should I do?▾
First, check the symlink status for that skill and agent in the Skills Desktop dashboard. If the status is 'Broken', it typically means the target agent's config directory path has changed (e.g., after a tool version update). Try triggering a 'Re-link' action within the app or remove and re-add the skill to rebuild the symlink. If the status shows 'Valid' but content is still outdated, some agents may cache skill content—simply restarting the AI agent tool usually resolves this.
View on GitHub

Project Metrics

Stars380
LanguageTypeScript
LicenseUnknown
Deploy DifficultyEasy

Table of Contents

  1. 01What is it?
  2. 02Pain Points vs Innovation
  3. 03Architecture Deep Dive
  4. 04Deployment Guide
  5. 05Use Cases
  6. 06Limitations & Gotchas
  7. 07Frequently Asked Questions

Related Projects

OpenClaw
OpenClaw
25.1 k·TypeScript
DeerFlow — ByteDance Open-Source SuperAgent Harness
DeerFlow — ByteDance Open-Source SuperAgent Harness
26.1 k·Python
gstack
gstack
0·TypeScript
Marketing for Founders
Marketing for Founders
2.2 k·Markdown