"screenshot" matching MCP servers:

Fork, break, and revert real VirtualBox VMs as MCP tools; drive GUI-only guests through the hypervisor (screenshot, click, type, keys) with nothing installed in the guest. Windows host with VirtualBox.

Visual feedback as agent work packets: stakeholders pin on your live app and your AI coding agent reads each pin (selector, screenshot, DOM, thread, acceptance criteria) via MCP and ships the fix.

Lets AI agents build, run and live-test FiveM servers by sending server and client console commands over RCON and devcon, driving the game window with keyboard, mouse and screenshot automation, and invoking in-game natives, exports and NUI callbacks through a companion bridge resource.

Enables headless batchmode control of Unity projects for compile, test, build, and screenshot tasks, avoiding the need for an interactive editor.

Enables MCP-based interaction with AutoCAD and DXF/DWG drawings, supporting document lifecycle, entities, layers, blocks, layouts, viewports, zoom, screenshot capture, and PDF export via ezdxf or live Windows AutoCAD COM backends.

An MCP server that enables users to control a full Chrome browser running in a Daytona cloud sandbox, allowing web browsing, interaction, and screenshot capture through Claude or other MCP clients.

Enables AI assistants to automate Wayland desktop environments through screenshot analysis, mouse control, and keyboard input simulation. It supports visual context via VLM providers like Gemini and OpenRouter to perform complex, multi-step desktop actions.

Enables AI assistants to perform browser automation using Playwright, including navigation, content extraction, screenshot analysis, and custom script execution.

Securely control your Mac from AI assistants using native mouse, keyboard, screenshot, and shell tools.

Provides local Model Context Protocol tools for invited device and My Apps collaboration, enabling terminal connection, app inspection, file and screenshot operations, and guided device control with human approval.