A desktop app for building, organizing, syncing, and backing up AI agent skills across tools, scenarios, and project workspaces.
Skills-Manager-Plus is a cross-tool skill manager for people who work with multiple AI coding agents. It gives you one central place to install skills, organize them by scenario, sync them to supported tools, compare them with project-local skill folders, and keep everything versioned with Git.
Instead of managing scattered skills/ directories by hand, you maintain a central library and decide how each workflow should use it.
- Skills are usually scattered across different tools and machines.
- Different workflows need different active skill sets.
- Project-local skills drift away from central libraries over time.
- Backing up or migrating a skills library is easy to forget and hard to recover cleanly.
Skills-Manager-Plus addresses this with a central library, scenario-based switching, project workspace comparison, per-agent sync, and built-in Git backup.
- Central skill library — Import skills from Git repositories, local folders,
.zipor.skillarchives, skills.sh, ClawHub, and plugin bundles. - Skills Management — Review, tag, enable, disable, inspect, compare, and sync skills from one management surface.
- Skills Store — Install from marketplace sources, Git repositories, local scans, ClawHub, and plugin markets.
- Scenarios — Keep separate skill sets for different workflows, clients, or projects, then switch between them quickly.
- Project Workspaces — Compare project-local skills against the central library and move changes in either direction.
- Per-agent sync — Sync with symlink or copy mode, including independent sync mode for custom agents.
- Git backup — Create snapshot-based history for backup, recovery, and multi-machine sync.
This fork extends the original skills-manager with deeper workflow coverage around discovery, local management, custom agents, and plugin distribution.
- Built-in ClawHub search and browsing flow.
- API key configuration in
Settings. - Integrated into the installation workflow instead of being a separate external step.
Skills ManagementseparatesSkill RepositoryandLocal Skills.- Local scan and import are part of the main management workflow.
- Batch import, source filtering, and description inference improve large local libraries.
- Each custom agent can use its own sync mode.
- Sync behavior can differ from the global default.
- Path and symlink handling are more robust during sync operations.
- Add plugin market sources from GitHub repositories.
- Discover plugin bundles and install packaged skills directly in-app.
- Track plugin-installed skills separately from the rest of the library.
- Separate bundle ID, updater channel, config directory, repo path, and database file.
- Can coexist with the original Skills Manager on the same machine.
The current app is organized around these modules:
DashboardSkills ManagementSkills StoreScenariosProject WorkspacesSettings
Detailed usage guides for each module live in docs/usage/README.md.
- Usage Overview
- Dashboard
- Skills Store
- Skills Management
- Scenarios
- Project Workspaces
- Settings
- Git Backup
Cursor · Claude Code · Codex · OpenCode · Amp · Kilo Code · Roo Code · Goose · Gemini CLI · GitHub Copilot · Windsurf · TRAE IDE · Antigravity · Clawdbot · Droid
You can also add custom tools in Settings.
- Create or switch to a scenario.
- Open
Skills Storeto import skills from marketplace, Git, local folders, ClawHub, or plugins. - Open
Skills Managementto decide which skills belong to the current scenario. - Sync enabled skills to installed tools.
- Use
Project Workspacesif you need to compare or exchange project-local skills. - Configure
Git Backupif you want history, restore points, or multi-machine sync.
- Node.js 18+
- Rust toolchain
- Tauri prerequisites for your OS
npm install
npm run tauri:devnpm run tauri:buildIf macOS blocks the downloaded app, run:
xattr -cr /Applications/Skills-Manager-Plus.appReplace the path if the .app file is stored somewhere else.
Skills-Manager-Plus is a fork of skills-manager v1.14.1.
Respect to the original project for building the foundation of cross-tool AI skill management.
MIT
