What problem does it solve?
AI agents often struggle to automate user-level GUI tasks when no public API exists, making simple actions like clicking, typing, or reading on-screen content time-consuming and error-prone.
Core Features & Use Cases
- It provides OS-wide desktop automation capabilities for Windows, macOS, and Linux; 42 tools in total, plus 6 compact tools when using MCP or REST interfaces; support for starting a built-in autonomous agent or exposing a tools server for external AI.
- It uses a Skill Unit concept defined by SKILL.md frontmatter (name and description) and can optionally reference scripts, references, and assets to extend behavior.
- Real-world use cases include GUI automation for data entry, form filling, file operations, window management, and browser automation in environments without public APIs.
Quick Start
Install the skill, grant necessary system permissions if required, and run clawdcursor start to begin autonomous desktop automation.