Skills Manager Desktop App (skills-manager)

How application developers use the RobOS Skills Manager (packages/skills-manager) to browse, parameterize, execute, and author 74+ system, Git, and Docker terminal skills directly from the desktop and within <robos-ai-textarea> prompt bars.

Table of contents

  1. Overview: Shell Skill Library for AI & Developers
  2. 10 Built-In Skill Categories
  3. Interactive Features
    1. 1. Instant Parameterization & Search
    2. 2. Custom User Skills
    3. 3. Community Skill Packs & Pattern Inspection
  4. Integration with <robos-ai-textarea>
  5. How to Launch and Test

Overview: Shell Skill Library for AI & Developers

While AI agent skills handle complex multi-step orchestration across the codebase, developers also need fast, deterministic shell commands for daily workstation operations: inspecting running Docker containers, finding large files, cleaning up merged Git branches, or profiling memory consumption.

The RobOS Skills Manager (packages/skills-manager) provides an interactive GUI skill catalog shipping with 74+ built-in shell skills across 10 operational categories:

RobOS Skills Manager Catalog Overview
RobOS Skills Manager Live Catalog: Interactive desktop grid displaying 61+ built-in workstation skills, category navigation pills, fast typeahead search, and inline parameter fields. (Click image to zoom full screen)
RobOS Skills Manager & Shell Architecture
RobOS Skills Manager & Shell Architecture: Interactive catalog of 74+ shell skills and custom macros accessible via GUI and prompt completion, executing deterministically in POSIX shell and Tilix terminal. (Click image to zoom full screen)

10 Built-In Skill Categories

The Skills Manager organizes commands into 10 structured categories:

Category Skill Count Common Operations & Commands
Git Operations 8 Recent commits (git log --oneline --graph), changed files (git status --short), branch lists (git branch -vv), merged branch cleanup, top contributors, large objects in history.
Docker / Containers 5 Container list (docker ps -a), live resource stats (docker stats --no-stream), image inspection, volume cleanup (docker system prune -f --volumes).
File Operations 7 Find large files (find ~ -size +100M), recent edits (find . -mtime -1), disk usage summary (du -sh */), count files by extension, find empty directories, find duplicates via MD5.
Process Management 7 Top memory consumers (ps aux --sort=-%mem), top CPU consumers, open ports (ss -tlnp), free memory & virtual memory stats, zombie processes, process tree (pstree).
Network Diagnostics 5 Internet connectivity test (curl -s https://github.com), network interfaces (ip addr), active TCP connections (ss -tp), DNS resolution (dig), cumulative bandwidth usage.
System Overview 6 Kernel & uptime (uname, uptime), CPU details (lscpu), OS release (lsb_release), running systemd services, failed system services, recent system errors (journalctl -p err).
Package Management 4 Installed global npm packages, outdated npm packages, pip packages list, recently installed apt packages.
Text Processing 4 Pretty-print JSON (jq), count & sort lines (sort | uniq -c), preview CSV files, recursive text grep.
Security & Auditing 5 SSH key inventory and fingerprints (ssh-keygen -lf), GPG key list, recent user login history (last), sudo usage audit, open file descriptors count.
Development Runtimes 4 Node & npm versions, Python version, current environment variables, port-in-use finder (ss -tlnp | grep :<port>).
Category Navigation — Git Operations
Category Filtering (Git Operations): Instantly focus on version control macros, commit log visualizers, branch cleanup, and contributor statistics with one-click terminal execution. (Click image to zoom full screen)

Interactive Features

1. Instant Parameterization & Search

Filter by category tab (File, Process, Git, Network, Docker, etc.) or type instant search terms. For commands requiring runtime parameters (such as $PORT, $FILE, $PATTERN, or $BRANCH), the UI automatically detects bash variables and renders dynamic inline input boxes so arguments can be customized prior to execution:

Dynamic Search and Inline Parameter Fields
Instant Search & Parameter Binding: Typing "port" instantly narrows the view to port-related utilities, rendering an editable input field for $PORT that seamlessly interpolates into ss -tlnp | grep :$PORT. (Click image to zoom full screen)

2. Custom User Skills

Developers can add personal bash commands and scripts directly from the UI by clicking + New Skill. Custom skills are saved to ~/.config/robos/skills.json and persist across reboots, instantly becoming available in the desktop grid and across all AI prompt bars.

Authoring Custom Skills Modal
Custom Skill Authoring Dialog: Declare custom shell macros with variable placeholders (e.g. git stash push -m "$MSG"), assigning categories and descriptions for immediate workstation availability. (Click image to zoom full screen)

3. Community Skill Packs & Pattern Inspection

Under the Skill Packs tab, developers can browse curated collections of skills and patterns tailored for specific engineering domains: SDLC Essentials, Cloud Native, Security Hardening, and Frontend Toolkit.

Skill Packs Marketplace
Skill Packs Marketplace: Browse, install, and manage curated skill bundles designed for Kubernetes, SDLC workflows, security auditing, and modern frontend frameworks. (Click image to zoom full screen)

Clicking View Patterns opens an in-depth pattern inspector showing the full library of prompt engineering templates, required parameters, and generated context instructions:

Pattern Browser & Prompt Preview
Pattern Browser & Live Prompt Preview: Inspect structured AI prompt templates (Code Review, Refactoring, Test Generation) with markdown formatting, parameter placeholders, and direct copy-to-clipboard functionality. (Click image to zoom full screen)

Integration with <robos-ai-textarea>

Every skill registered in the Skills Manager is automatically wired into the <robos-ai-textarea> web component used across RobOS applications:

  • Type @ in any AI textarea to trigger typeahead.
  • Select @skill:top-memory or @skill:git-changed-files.
  • The AI agent automatically executes the diagnostic command and incorporates the terminal output into its planning or prompt context.

How to Launch and Test 

# Launch the Skills Manager (or open "Skills Manager" in RobOS App Launcher):
electron packages/skills-manager

# Run the comprehensive End-to-End test suite with screenshot capture:
node --test packages/robos-test/tests/skills-manager/e2e.test.js

# Run the automated smoke test suite:
node --test packages/robos-test/tests/skills-manager/smoke.test.js

# Execute the narrated E2E video demo walkthrough:
node packages/robos-test/demos/skills-manager-demo.js