tterm

tterm is a macOS cockpit specifically designed for developers to streamline their workflow when using Claude Code. It consolidates essential development tools into a unified interface, allowing users to manage multiple projects simultaneously. Each project is presented as a row containing an explorer, the Claude Code interface, and a file viewer, enabling efficient navigation and management of concurrent tasks. The core problem tterm addresses is the fragmentation of developer workflows, which often requires juggling multiple windows for different tools. Developers typically need a terminal for running code, a browser to preview and test their work, and an editor for writing code. This constant switching between applications can lead to reduced productivity and increased cognitive load. tterm aims to solve this by bringing these critical components together into a single, cohesive environment. One of tterm's key features is its integrated file explorer, which provides a clear view of project files within each project row. This allows developers to easily navigate their project structure without leaving the tterm application. Complementing this is the integrated Claude Code pane, which serves as the AI coding assistant. Users can interact with Claude Code directly within tterm to generate code, refactor existing code, or ask for assistance with development tasks. A significant capability of tterm is its built-in Chromium browser. This real browser instance allows developers to preview their web applications and test changes in real-time, directly within the tterm interface. The browser syncs with the user's Chrome bookmarks and cookies, providing a familiar browsing environment. Notably, tterm intentionally omits a traditional code editor, encouraging developers to review changes via diffs and commit them incrementally, fostering a more deliberate development process. The self-rebuilding functionality is a unique aspect of tterm. Users can ask the Claude pane to implement a new feature, and the running application will hot-reload with the changes. This dynamic rebuilding process allows for rapid iteration and feature development directly within the cockpit. This feature aims to make the development of the tool itself as fluid as the development it facilitates. tterm operates on a methodology that integrates AI assistance directly into the development lifecycle. By placing Claude Code alongside the file explorer and a live browser preview, it creates a feedback loop where code changes can be immediately tested and refined. The emphasis on reviewing diffs and committing incrementally, combined with the hot-reloading feature, promotes an agile and responsive development approach. The benefits for users include increased productivity through a consolidated workflow, reduced context switching, and faster iteration cycles thanks to the integrated AI and hot-reloading capabilities. The absence of an editor encourages a focus on code review and version control best practices. Furthermore, the commitment to no account and no telemetry ensures user privacy and a distraction-free experience. Concrete use cases for tterm include rapidly prototyping new features by describing them to Claude Code and seeing them appear in the browser preview, debugging issues by reviewing code changes and testing them instantly, and managing multiple development projects side-by-side within a single application window. It's ideal for developers who want to leverage AI for coding tasks without leaving their primary development environment. tterm is available for macOS on Apple silicon and is free for hobbyists. It does not require an account and collects no telemetry, emphasizing user privacy and simplicity. The application integrates with the user's Chrome bookmarks and cookies, with explicit consent prompts for accessing sensitive data like decryption keys from the macOS Keychain. In essence, tterm offers a novel, integrated environment for developers to harness the power of AI, streamline their coding and testing processes, and build applications more efficiently on macOS.