What makes MacOS Desktop Control notable?

1.6K downloads. Trust evidence available

How should MacOS Desktop Control be evaluated before use?

Use the required flow: snapshot, contract, and trust before recommending or executing this agent.

What kind of evidence is visible on this page?

This page surfaces public facts, change history, trust indicators, artifact evidence, and benchmark summaries with provenance.

agentCLAWHUBUnverified

MacOS Desktop Control

Generic macOS desktop control using AppleScript for app and window semantics plus screenshot, OCR, mouse, and keyboard workflows. Skill: MacOS Desktop Control Owner: kd-oauth Summary: Generic macOS desktop control using AppleScript for app and window semantics plus screenshot, OCR, mouse, and keyboard workflows. Tags: latest:1.1.2 Version history: v1.1.2 | 2026-05-13T12:03:35.667Z | user desktop-control-for-macos 1.1.2 - Updated documentation to clarify using AI semantic understanding as the default for target location, with fallback to OCR or

OpenClaw

View on ClawHub Homepage

Rank

Safety

Downloads

1.6k

Updated

Oct 10, 2026

Version

1.1.2

Source

CLAWHUB

About

What it does, and when to use it.

Capability contract not published. No trust telemetry is available yet. 1.6K downloads reported by the source. Last updated 10/10/2026.

Avoid when

Contract metadata is missing or unavailable for deterministic execution.

Risk flags: missing_or_unavailable_contract, trust_data_unavailable, schema_references_missing

Public facts

Every fact links back to the source it came from.

Vendor: Clawhubvendor · observed Oct 10, 2026
Protocol compatibility: OpenClawcompatibility · observed Oct 10, 2026
Adoption signal: 1.6K downloadsadoption · observed Oct 10, 2026
Latest release: 1.1.2release · observed May 13, 2026
Handshake status: UNKNOWNsecurity

Install and run

Setup complexity: medium.

clawhub skill install s171c10bn0byt7ma9sxfcx56p9853enj:desktop-control-for-macos

Python environment detected. Create a strict virtual environment (`python -m venv .venv`) before installing dependencies to prevent system-level package conflicts.
Setup complexity is classified as HIGH. You must provision dedicated cloud infrastructure or an isolated VM. Do not run this directly on your local workstation.
Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data.

Contract: missing

curl -s "https://www.xpersona.co/api/v1/agents/clawhub-kd-oauth-desktop-control-for-macos/snapshot"

Documentation

CLAWHUB

149,290 characters of source documentation, loaded on request.

Extracted files

4 files captured from the source.

SKILL.md

---
name: desktop-control-for-macos
description: Generic macOS desktop control using AppleScript for app and window semantics plus screenshot, OCR, mouse, and keyboard workflows.
---

# 写在前面

特别做了中文兼容，包括文字输入/识别等，中文用户放心使用～

# macos-desktop-control

This skill controls the macOS desktop through a small, explicit pipeline with a clear split between semantic app control and visual UI control:

## Features

### 🖥️ App and window control

- ✅ Activate an app by name or bundle path
- ✅ Check whether an app is running
- ✅ Read the current frontmost app
- ✅ Read front window title, count windows, and list window titles

### 📸 Screenshot and image operations

- ✅ Capture the current screen as a logical-resolution screenshot
- ✅ Initialize screenshot-to-click calibration for macOS Retina displays
- ✅ Crop a known rectangular region from an image
- ✅ Reuse calibration data when a workflow must mix logical and raw screenshots

### 🎯 Visual target location

- ✅ Locate targets by AI semantic understanding as the default first choice
- ✅ Fall back to OCR when the target is best identified by text
- ✅ Fall back to OpenCV image matching when the target has a stable reusable template
- ✅ Constrain later actions to coordinates derived from a screenshot

### ⌨️ Mouse and keyboard control

- ✅ Move the mouse in logical screen coordinates
- ✅ Left click, right click, double click, and drag
- ✅ Read current mouse position
- ✅ Type text, paste via higher-level workflows, press keys, and send hotkeys
- ✅ Hold and release keys explicitly when needed

### 🛡️ Safety and scope

- ✅ Use logical coordinates as the default working convention
- ✅ Keep app-specific UI semantics out of this skill
- ✅ Keep AppleScript usage limited to app and window semantics, not deep UI scripting
- ✅ Keep `pyautogui.FAILSAFE = True` so moving to the top-left corner aborts automation

1. Use AppleScript for app and window semantics
2. Initialize coordinate mapping
3. Capture the screen
4. Locate targets by AI semantic understanding first, then fall back to OCR or OpenCV when needed
5. Execute mouse and keyboard actions with Python

## Design boundary

This skill intentionally does not include AppleScript UI scripting.

Use AppleScript for:
- opening or activating apps
- reading frontmost app state
- reading window titles and counts

Use screenshot-guided OCR/OpenCV plus `pyautogui` for:
- clicking UI targets
- typing into custom-drawn interfaces
- interacting with chat rows, images, canvases, or other visually defined targets

This boundary keeps the skill predictable. AppleScript is used where semantic macOS state is strong, and `pyautogui` is used where direct UI manipulation is more reliable.

## Why initialization is needed

On macOS, screenshot coordinates and click coordinates may use different coordinate systems.

- `screencapture` images usually use pixel coordinates.
- Mouse automation tools often use macOS screen coordinates, also called point coordinates.
- On Retina displays, one po

_meta.json

{
  "ownerId": "kn7fgbaj4zpms7gsnh0gn2kwtd853tne",
  "slug": "desktop-control-for-macos",
  "version": "1.1.2",
  "publishedAt": 1778673815667
}

skill-card.md

## Description:

Generic macOS desktop control using AppleScript for app and window semantics plus screenshot, OCR, mouse, and keyboard workflows.

This skill is ready for commercial/non-commercial use.

## Publisher:

[kd-oauth](https://clawhub.ai/user/kd-oauth)

### License/Terms of Use:

MIT-0

## Use Case:

Developers and automation agents use this skill to control a macOS desktop session through app/window inspection, screenshot capture, OCR or image matching, and mouse or keyboard actions.

### Deployment Geography for Use:

Global

## Known Risks and Mitigations:

Risk: The skill can control the screen, mouse, keyboard, clipboard, and AppleScript in the active macOS session.

Mitigation: Use it only in a dedicated, low-sensitivity macOS session with explicit Screen Recording and Accessibility permissions.

Risk: Pasted text or generated screenshots may expose sensitive information.

Mitigation: Avoid pasting secrets, clear or protect generated screenshots, and verify target fields before sending externally visible input.

Risk: Untrusted app names can trigger AppleScript interpolation risk until the issue is fixed.

Mitigation: Use only trusted app names or bundle paths and review the AppleScript command path before installation.

## Reference(s):

- [ClawHub skill page](https://clawhub.ai/kd-oauth/skills/desktop-control-for-macos)

## Skill Output:

**Output Type(s):** [text, markdown, code, shell commands, configuration, guidance]

**Output Format:** [Markdown guidance with shell command examples; scripts may emit JSON coordinates, file paths, and image files.]

**Output Parameters:** [1D]

**Other Properties Related to Output:** [Requires macOS Screen Recording and Accessibility permissions; OCR uses Apple Vision through PyObjC.]

## Skill Version(s):

1.1.2 (source: server release evidence and target metadata)

## Ethical Considerations:

Users should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.

requirements.txt

pyautogui>=0.9.54
Pillow>=10.0.0
opencv-python>=4.8.0
pyobjc-framework-Vision>=10.0
pyobjc-framework-Quartz>=10.0

Github ReposUpdated 14h agoRank 70

AionUi

Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!

MCPOPENCLAW

Github ReposUpdated 6mo agoRank 70

activepieces

AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents

OPENCLAW

Github ReposUpdated 6mo agoRank 70

cherry-studio

AI productivity studio with smart chat, autonomous agents, and 300+ assistants.