Video Still Capture MCP is a Python-based server implementing the Model Context Protocol (MCP) that enables AI assistants to access and control webcams and OpenCV-compatible video sources. It exposes a compact set of tools that let language models capture images, adjust camera properties, and manage video connections in a simple, scriptable API. Core capabilities include quick image capture, opening and maintaining camera connections, capturing frames from a given connection, and closing connections to release resources. The API also exposes get_video_properties to read properties such as width, height, and fps, and set_video_property to modify values like brightness, contrast, and resolution. A basic image processing feature supports transformations like horizontal flipping. The server integrates with Claude Desktop installations via UV, enabling seamless control through natural language prompts, and can be used with multiple cameras by specifying device indices. Resource management is emphasized to ensure cameras are released on shutdown and typical workflows involve opening connections, capturing frames, adjusting settings, and closing connections when done.