Kira Mcp
About
A local Model Context Protocol (MCP) server that gives any MCP-compatible host (Claude Desktop, Claude Code, Cursor, Cline, Continue) full computer-use control of the machine. One perceive_screen call screenshots the display, runs a bundled OmniParser-v2 YOLO icon-detector plus o
Details
- Author
- Anmol202005
- Downloads
- 265
- Categories
- Developer Tools, Automation
Jump to
- Screenshots the display and runs OmniParser-v2 YOLO icon detection.
- Optional OCR for text recognition on screen.
- Returns click-ready (cx, cy) pixel coordinates.
- Full mouse, keyboard, and clipboard automation via pyautogui.
- No API key required – fully local processing.
- Model weights are shipped inside the package.
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Kira McpCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Install the package with pip install kira-mcp, then configure the server in any MCP-compatible host (e.g., Claude Desktop, Claude Code, Cursor, Cline, Continue). The server exposes the perceive_screen tool, which captures the screen and returns actionable coordinates for automation.
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"kira mcp": {
"kira-mcp": {
"command": "kira-mcp"
}
}
}
}McpServers
{
"kira-mcp": {
"command": "kira-mcp"
}
}A local Model Context Protocol (MCP) server that gives any MCP-compatible host (Claude Desktop, Claude Code, Cursor, Cline, Continue) full computer-use control of the machine. One perceive_screen call screenshots the display, runs a bundled OmniParser-v2 YOLO icon-detector plus optional OCR, and returns click-ready (cx, cy) pixel coordinates, with full mouse, keyboard, and clipboard automation via pyautogui. No API key, no network call. the detector weights ship inside the wheel. Install with: pip install kira-mcp
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
