What is borg?

Borg is a small AI coding assistant that lives in your terminal. You give it a task in plain English, and it reads your files, makes edits, runs commands, and keeps checking its own work against your project's build or test setup until the task is actually finished, rather than just guessing once and stopping. What sets it apart is how it connects to an AI model. You can point it at a model running entirely on your own computer through tools like Ollama or LM Studio, at other services such as OpenAI or OpenRouter, or you can log into the maker's own hosted service, xShellz, and use their models instead through a metered plan. The agent itself, its tools, and the way it works stay identical no matter which option you pick, so switching backends is just a setting change. The tool ships as a single small program with no separate install steps for a runtime or extra dependencies, and it uses very little memory while idle. It asks for permission before it is allowed to write files inside a folder, and any actions like editing or running shell commands happen right there on your own machine. Only the actual thinking, the AI model call itself, goes out to wherever that model is hosted. If you use your own local model, no code or key ever leaves your computer, and any provider key you do use is read from your environment rather than being written to a settings file. Inside its interactive mode you can switch models, ask it to study a project and write itself some notes about it, resume earlier conversations, and check how much of its available context window is being used. It works with tool-calling capable local models such as Qwen 2.5 Coder, or with larger hosted models for more difficult tasks. The project is open source under the Apache 2.0 license, which allows free use, modification, and distribution, including commercially, as long as you keep the license and copyright notices and note any changes you make.