AZMX AI
A terminal-based AI coding agent platform, distributed as a small native installer, that talks directly to a chosen provider or a local Ollama model, asks approval before writing files or running commands, and bundles a terminal, editor and file explorer.
AZMX AI is a terminal-first AI agent application distributed as a small, roughly 10 MB, native installer for macOS, Linux and Windows, with no Electron runtime and a cold start under a second. It never proxies AI requests through its own servers: prompts go directly from the user's device to whichever provider is chosen, or to a local model.
It offers one-click local setup with Ollama and curated coding models such as Qwen2.5-Coder, Granite Code and Llama, or bring-your-own-key access to OpenAI, Anthropic, Google, Groq, xAI, Cerebras, DeepSeek, NVIDIA NIM, Azure OpenAI or any OpenAI-compatible endpoint. Every agent action that writes a file or runs a shell command asks for approval first, with a configurable strictness level, and every tool call is recorded in a hash-chained, tamper-evident local audit log that can be exported on paid tiers.
It bundles a real terminal, a code editor with vim mode and inline AI autocomplete, a file tree with git status badges, and a curated catalog of 17 MCP servers that can be extended with a single JSON file. The application itself is proprietary; this repository publishes release artifacts, the auto-updater manifest, documentation and community contributions such as skills and agents.