A high-fidelity automation bridge for macOS (Darwin) that enables agents to perceive the desktop state and execute precise mouse and keyboard interactions.
🛠 Included Scripts
This skill leverages specialized wrappers located in the /scripts directory to interface with system-level binaries:
1. cliclick_wrapper.sh
A dedicated wrapper for the cliclick utility to handle synthetic input events.
Path: scripts/cliclick_wrapper.sh
Logic: Executes /opt/homebrew/bin/cliclick with passed arguments.
Capabilities: Left/Right clicks, mouse movement, and keyboard emulation.
2. vision_wrapper.sh
The visual perception engine for the skill.
Path: scripts/vision_wrapper.sh
Logic: Utilizes the native macOS screencapture utility in silent mode (-x).
Output: Generates a standard PNG at /tmp/claw_view.png.
🚀 Tool Specifications
see
Captures the current screen state for visual analysis.
Returns: A string confirming the filepath of the capture.
Use Case: Identifies UI elements, window positions, and application states.
click
Sends precise commands to the mouse and keyboard.
Usage: click "c:x,y" (Click) or click "m:x,y" (Move).
Syntax: Supports all cliclick standard notation including w: (wait) and t: (type).
⚙️ Requirements & Setup
Binary Dependency:
brew install cliclick
1---2name: macos-desktop-control3description: macOS Control Skill4---5# macOS Control Skill67A high-fidelity automation bridge for macOS (Darwin) that enables agents to perceive the desktop state and execute precise mouse and keyboard interactions.89## 🛠 Included Scripts10This skill leverages specialized wrappers located in the `/scripts` directory to interface with system-level binaries:1112### 1. `cliclick_wrapper.sh`13A dedicated wrapper for the `cliclick` utility to handle synthetic input events.14- **Path**: `scripts/cliclick_wrapper.sh`15- **Logic**: Executes `/opt/homebrew/bin/cliclick` with passed arguments.16- **Capabilities**: Left/Right clicks, mouse movement, and keyboard emulation.1718### 2. `vision_wrapper.sh`19The visual perception engine for the skill.20- **Path**: `scripts/vision_wrapper.sh`21- **Logic**: Utilizes the native macOS `screencapture` utility in silent mode (`-x`).22- **Output**: Generates a standard PNG at `/tmp/claw_view.png`.2324---2526## 🚀 Tool Specifications2728### `see`29Captures the current screen state for visual analysis.30* **Returns**: A string confirming the filepath of the capture.31* **Use Case**: Identifies UI elements, window positions, and application states.3233### `click`34Sends precise commands to the mouse and keyboard.35* **Usage**: `click "c:x,y"` (Click) or `click "m:x,y"` (Move).36* **Syntax**: Supports all `cliclick` standard notation including `w:` (wait) and `t:` (type).3738---3940## ⚙️ Requirements & Setup41421. **Binary Dependency**:43 ```bash44 brew install cliclick45 ```
Run npx skillmds@latest add javimosch/macos-desktop-control in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.
macOS Control Skill It is listed under Coding & Dev Tools on SkillMD.
This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.
This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.
Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.
javimosch (@javimosch) published this skill. Their other Agent Skills are listed on their SkillMD profile.
javimosch