Hey @Vashti K ! Happy to break it down. At its core, Bliss and the Chorus engine are built to solve a big problem with standard AI chat windows: traditional interfaces don't know anything about your local code or tools. Bliss acts as a web-based "desktop" environment, while the Chorus engine works under the hood as a custom development harness. It lets you interact with AI models while keeping them fully aware of your local files, codebases, and project context. Making it a real workspace rather than just a simple text prompt and hold thumbs that it worked. Under the hood, it operates as a hybrid engine designed for speed and precision. It uses SQLite-VEC to turn your local project files into a searchable vector database, so the AI can pull exact code snippets in milliseconds without needing expensive cloud databases. On top of that, I built a high-performance Rust subkernel (HIVE) and a custom language (ChorusDSL) that handles asynchronous execution and tracks state safely. It can seamlessly hook into both local models on your machine and heavy-hitter cloud endpoints like Gemini (Anthropic API endpoints might be added at some point). What makes it great is the speed, context, and developer sovereignty. Instead of copying and pasting code back and forth or locking yourself into locked-down cloud platforms, Chorus gives you a fast, modular dev tool that runs right where your code lives. Plus, the whole workflow is continuous. You prototype features in the workspace, refine them through the local agent harnesses, push to wherever, get a 2nd opinion from a cloud hosted beefer or just deploy on your local stack.