Computer Use

Community

viralcode

computer-use

Full desktop computer use for headless Linux servers and VPS. Creates a virtual display (Xvfb + XFCE) to control GUI applications without a physical monitor. Screenshots, mouse clicks, keyboard input, scrolling, dragging — all 17 standard actions. Includes flicker-free VNC setup for live remote viewing. Model-agnostic, works with any LLM.

Overview

Publisher	viralcode
Repository	`openwhale`
Skill name	`computer-use`
Stars	65
Forks	11
Bundled files	18
Links	Skill source Repository Homepage

Markdown instructions
A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.
Works with any LLM
AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.
18 bundled files
Scripts, templates, and references the model can read while it works. Files are read-only and never executed.
Open source
Published by viralcode on GitHub. Read the source before you install it.

Installation

Install the Computer Use AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

Open the app and go to Plugins → Skills.
Choose "Install from GitHub".
Paste the skill folder URL below and confirm.
Enable the skill in any chat where you want it available.

Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills

git clone --depth 1 https://github.com/viralcode/openwhale.git /tmp/openwhale
mkdir -p .claude/skills
cp -r /tmp/openwhale/skills/computer-use .claude/skills/computer-use

Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Computer Use in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Computer Use on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Computer Use is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

Loaded only when it is needed
The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.
Switch models mid-chat
Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Computer Use Skill

Full desktop GUI control for headless Linux servers. Creates a virtual display (Xvfb + XFCE) so you can run and control desktop applications on VPS/cloud instances without a physical monitor.

Environment

Display: :99
Resolution: 1024x768 (XGA, Anthropic recommended)
Desktop: XFCE4 (minimal — xfwm4 + panel only)

Quick Setup

Run the setup script to install everything (systemd services, flicker-free VNC):

bash
./scripts/setup-vnc.sh

This installs:

Xvfb virtual display on :99
Minimal XFCE desktop (xfwm4 + panel, no xfdesktop)
x11vnc with stability flags
noVNC for browser access

All services auto-start on boot and auto-restart on crash.

Actions Reference

Action	Script	Arguments	Description
screenshot	`screenshot.sh`	—	Capture screen → base64 PNG
cursor_position	`cursor_position.sh`	—	Get current mouse X,Y
mouse_move	`mouse_move.sh`	x y	Move mouse to coordinates
left_click	`click.sh`	x y left	Left click at coordinates
right_click	`click.sh`	x y right	Right click
middle_click	`click.sh`	x y middle	Middle click
double_click	`click.sh`	x y double	Double click
triple_click	`click.sh`	x y triple	Triple click (select line)
left_click_drag	`drag.sh`	x1 y1 x2 y2	Drag from start to end
left_mouse_down	`mouse_down.sh`	—	Press mouse button
left_mouse_up	`mouse_up.sh`	—	Release mouse button
type	`type_text.sh`	"text"	Type text (50 char chunks, 12ms delay)
key	`key.sh`	"combo"	Press key (Return, ctrl+c, alt+F4)
hold_key	`hold_key.sh`	"key" secs	Hold key for duration
scroll	`scroll.sh`	dir amt [x y]	Scroll up/down/left/right
wait	`wait.sh`	seconds	Wait then screenshot
zoom	`zoom.sh`	x1 y1 x2 y2	Cropped region screenshot

Usage Examples

bash
export DISPLAY=:99

# Take screenshot
./scripts/screenshot.sh

# Click at coordinates
./scripts/click.sh 512 384 left

# Type text
./scripts/type_text.sh "Hello world"

# Press key combo
./scripts/key.sh "ctrl+s"

# Scroll down
./scripts/scroll.sh down 5

Workflow Pattern

Screenshot — Always start by seeing the screen
Analyze — Identify UI elements and coordinates
Act — Click, type, scroll
Screenshot — Verify result
Repeat

Tips

Screen is 1024x768, origin (0,0) at top-left
Click to focus before typing in text fields
Use ctrl+End to jump to page bottom in browsers
Most actions auto-screenshot after 2 sec delay
Long text is chunked (50 chars) with 12ms keystroke delay

Live Desktop Viewing (VNC)

Watch the desktop in real-time via browser or VNC client.

Connect via Browser

bash
# SSH tunnel (run on your local machine)
ssh -L 6080:localhost:6080 your-server

# Open in browser
http://localhost:6080/vnc.html

Connect via VNC Client

bash
# SSH tunnel
ssh -L 5900:localhost:5900 your-server

# Connect VNC client to localhost:5900

SSH Config (recommended)

Add to ~/.ssh/config for automatic tunneling:


Host your-server
  HostName your.server.ip
  User your-user
  LocalForward 6080 127.0.0.1:6080
  LocalForward 5900 127.0.0.1:5900

Then just ssh your-server and VNC is available.

System Services

bash
# Check status
systemctl status xvfb xfce-minimal x11vnc novnc

# Restart if needed
sudo systemctl restart xvfb xfce-minimal x11vnc novnc

Service Chain


xvfb → xfce-minimal → x11vnc → novnc

xvfb: Virtual display :99 (1024x768x24)
xfce-minimal: Watchdog that runs xfwm4+panel, kills xfdesktop
x11vnc: VNC server with -noxdamage for stability
novnc: WebSocket proxy with heartbeat for connection stability

Opening Applications

bash
export DISPLAY=:99
google-chrome --no-sandbox &    # Chrome (recommended)
xfce4-terminal &                # Terminal
thunar &                        # File manager

Note: Snap browsers (Firefox, Chromium) have sandbox issues on headless servers. Use Chrome .deb instead:

bash
wget https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb
sudo dpkg -i google-chrome-stable_current_amd64.deb
sudo apt-get install -f

Manual Setup

If you prefer manual setup instead of setup-vnc.sh:

bash
# Install packages
sudo apt install -y xvfb xfce4 xfce4-terminal xdotool scrot imagemagick dbus-x11 x11vnc novnc websockify

# Copy service files
sudo cp systemd/*.service /etc/systemd/system/

# Edit xfce-minimal.service: replace %USER% and %SCRIPT_PATH%
sudo nano /etc/systemd/system/xfce-minimal.service

# Mask xfdesktop (prevents VNC flickering)
sudo mv /usr/bin/xfdesktop /usr/bin/xfdesktop.real
echo -e '#!/bin/bash\nexit 0' | sudo tee /usr/bin/xfdesktop
sudo chmod +x /usr/bin/xfdesktop

# Enable and start
sudo systemctl daemon-reload
sudo systemctl enable --now xvfb xfce-minimal x11vnc novnc

Troubleshooting

VNC shows black screen

Check if xfwm4 is running: pgrep xfwm4
Restart desktop: sudo systemctl restart xfce-minimal

VNC flickering/flashing

Ensure xfdesktop is masked (check /usr/bin/xfdesktop)
xfdesktop causes flicker due to clear→draw cycles on Xvfb

VNC disconnects frequently

Check noVNC has --heartbeat 30 flag
Check x11vnc has -noxdamage flag

x11vnc crashes (SIGSEGV)

Add -noxdamage -noxfixes flags
The DAMAGE extension causes crashes on Xvfb

Requirements

Installed by setup-vnc.sh:

bash
xvfb xfce4 xfce4-terminal xdotool scrot imagemagick dbus-x11 x11vnc novnc websockify

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Computer Use AI skill do?

Why use Computer Use on TypingMind?

Because you install it once and use it with any model. Computer Use is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Computer Use in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/viralcode/openwhale/tree/main/skills/computer-use. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Computer Use?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Computer Use?

As many as you like. As long as a model supports skills, you can use Computer Use with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Computer Use AI skill free?

It is published on GitHub by viralcode. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Executing Plans

obra

CommunityPopular

Use when you have a written implementation plan to execute in a separate session with review checkpoints

Finishing A Development Branch

obra

CommunityPopular

Use when implementation is complete, all tests pass, and you need to decide how to integrate the work

Receiving Code Review

obra

CommunityPopular

Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative…

Requesting Code Review

obra

CommunityPopular

Use when completing tasks, implementing major features, or before merging to verify work meets requirements

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

Computer Use Skill

Environment

Quick Setup

Actions Reference

Usage Examples

Workflow Pattern

Tips

Live Desktop Viewing (VNC)

Connect via Browser

Connect via VNC Client

SSH Config (recommended)

System Services

Service Chain

Opening Applications

Manual Setup

Troubleshooting

VNC shows black screen

VNC flickering/flashing

VNC disconnects frequently

x11vnc crashes (SIGSEGV)

Requirements

Bundled files

Frequently asked questions

Related AI Skills

Set up your own AI workspace now