Everywhere

Why we included this project

The pitch is simple: the assistant works from whatever is on your screen, so you are not shuffling context into a separate chat window. Everywhere watches the app you have open, picks a suitable LLM for the request, and uses MCP tools to reach into external services, which means it can carry out an action rather than just return text. That is genuinely useful if you spend your day moving between a code editor, browser, spreadsheets, and chat apps and are tired of re-explaining what you are working on. It ships as a cross-platform desktop app built on Avalonia, with documented configuration for wiring up your own models and tools and an active community behind it. If you are curious whether on-screen context can make an AI assistant feel less like a separate tool and more like part of the flow, this is a working example worth running against your own workflow.