Foxmayn AI CLI

Developers who live in the terminal had to break their workflow every time they needed to interact with an AI assistant. Web-based chat interfaces (ChatGPT, Claude, etc.) require opening a browser, navigating to the app, and context-switching away from the code they're working on. For quick questions, code reviews, or brainstorming, this overhead adds up — and copy-pasting between terminal and browser strips formatting and context.

Build a full-featured terminal chat application that brings multi-model AI conversations directly into the developer's existing workflow — with proper Markdown rendering, streaming responses, and no browser dependency.

• Built an alt-screen TUI application in Go using Bubble Tea (Elm architecture) with a scrollable viewport, text input, and real-time streaming display.

• Integrated OpenRouter's API for multi-model support: 7 models including Google Gemini, DeepSeek, Grok, and OpenAI variants — switchable via an in-app /model picker.

• Implemented streaming response rendering with full Markdown support: code blocks, bold, italic, lists, and headers — all formatted with Lip Gloss directly in the terminal.

• Added slash-command autocomplete (/model, /clear, /quit) for quick actions without leaving the conversation flow.

• Configured hot-reload development via watchexec and simple .env-based API key management.

Developers can now chat with multiple AI models without leaving the terminal. Streaming Markdown rendering, in-app model switching, and slash commands keep the entire AI workflow within the same context where code is being written.