"author:SharonBruno62" matching MCP servers:

Lets an AI agent drive live mobile and desktop sessions of a web page and receive measured UI findings — contrast ratios, touch-target sizes, overflow and theme drift — as actionable text tied to specific selectors, rather than guessing from screenshots. A shared panel keeps the human and agent on the same live view, where pinned notes and frames can be read back by the agent.

Enables LLMs to work with local Python environments by listing environments and installed packages, installing dependencies, and running code or scripts. Also supports reading, writing, and listing files within a sandboxed working directory.