tterm

Product Introduction

1. Definition: tterm is a native macOS application (a developer cockpit or integrated development environment) that combines a terminal emulator (pty), a fully-featured Chromium browser instance, and a dedicated pane for Claude Code into a single, unified window.
2. Core Value Proposition: tterm exists to eliminate context-switching for developers by consolidating the core tools of modern development—the command line, a web browser for research and testing, and an AI coding assistant—into a single, persistent workspace. Its primary value is enabling a frictionless developer workflow where projects can be managed in parallel rows and the application itself can be modified via self-hosting and hot-reloading through its integrated AI.

Main Features

1. Project Row Workspace: Each open project or repository is organized into a horizontal row containing three core panes: a file system explorer, a Claude Code interface for AI-powered code generation and review, and a file viewer/editor. Users can stack multiple project rows vertically to manage concurrent tasks, such as working on a backend service and its frontend client simultaneously.
2. Integrated Chromium Browser (⌘B): tterm embeds a real, unbranded Chromium instance, not a simple WebView. This integrated browser shares your default Chrome/Chromium profile, importing cookies, bookmarks, history, and favicons. It functions as a fully-capable browser tab within the development environment, ideal for API documentation, testing web applications locally, or debugging.
3. AI-Powered Self-Modification & Hot-Reload: A key innovation is the ability to use the built-in Claude Code pane to modify tterm's own source code. By pressing ⌘K and instructing Claude to add a feature or fix a bug for "tterm itself," the application can hot-reload the new changes while running. This demonstrates a self-hosting development model where the tool builds and extends itself.
4. Git-Centric Diff Review (⌘G): tterm intentionally omits a traditional text editor for writing code from scratch. Instead, it focuses on the review process. The ⌘G command opens a diff review interface where changes generated by Claude Code are presented hunk by hunk. Users can visually inspect, stage, and commit code directly, promoting a review-and-commit workflow over manual retyping.
5. Non-Intrusive Keyboard Protocol: The application respects the sanctity of the terminal. It only intercepts keyboard shortcuts (chords) that use the Command (⌘) key. All Control (⌃) key combinations pass directly through to the underlying terminal process (pty), ensuring compatibility with command-line tools like tmux, vim, or emacs without interference.

Problems Solved

1. Pain Point: Developer context-switching fatigue from constantly alt-tabbing between a terminal, an IDE, a web browser, and an AI assistant window. This fragmentation breaks focus and reduces productivity.
2. Target Audience: The primary user personas are macOS developers on Apple Silicon machines, particularly full-stack developers, indie hackers, startup engineers, and hobbyists who use Claude Code extensively. It caters to those who manage multiple repositories or projects concurrently and value a streamlined, keyboard-driven workflow.
3. Use Cases:
   • Rapid Prototyping: Using Claude Code in one pane to generate code while instantly testing the output in the integrated browser and terminal in adjacent panes.
   • Multi-Repository Management: Working on a microservices architecture where each service lives in its own stacked project row within tterm.
   • Code Review & Refactoring: Leveraging the AI assistant to suggest improvements and using the built-in diff tool to review and commit changes efficiently.
   • Learning & Exploration: Keeping documentation open in the Chromium pane while following along with terminal commands and code examples in a unified view.

Unique Advantages

1. Differentiation: Unlike traditional IDEs (VS Code, JetBrains) or terminal multiplexers (tmux with panes), tterm natively and seamlessly integrates a production-grade browser and a first-class AI coding assistant as core, equally-important components. Unlike using a browser extension for AI, tterm provides a dedicated, persistent Claude Code pane with deep workflow integration.
2. Key Innovation: The self-modifying, hot-reloading architecture is its most distinctive technical feature. The capability to have the running application modify its own source code based on natural language requests to its integrated AI and then instantly apply those changes is a novel approach to tool development, blurring the line between using a tool and building it.

Frequently Asked Questions (FAQ)

1. Is tterm a terminal emulator or an IDE? tterm is best described as a developer cockpit or integrated workflow environment. It contains a terminal emulator (pty) but expands into a holistic workspace by integrating a browser and AI assistant, going beyond the scope of a traditional IDE or simple terminal.
2. How does tterm's integrated browser differ from a browser tab? tterm uses a real Chromium Embedded Framework (CEF) instance that loads your actual Chrome user profile. This means it has access to your authenticated sessions, extensions (that work with CEF), and bookmarks, making it a true replacement for a standalone browser window during development.
3. Can I use tterm with GitHub Copilot or other AI assistants besides Claude Code? No, tterm is specifically and exclusively designed to integrate with Anthropic's Claude Code. It is a core, non-replaceable component of the application's workflow. Using tterm requires a Claude Code subscription.
4. Is tterm safe to download and run on my Mac? Yes, the tterm.app is notarized by Apple, meaning it has passed Apple's security checks for malicious content. You can download it directly from the official GitHub releases, unzip it, and run it without disabling macOS security settings, as it is notarized.
5. Why does tterm only support Apple Silicon Macs (M1/M2/M3)? The application is likely built using native ARM64 binaries and frameworks to leverage the performance and efficiency of Apple Silicon. Supporting Intel Macs would require maintaining a separate x86_64 build and may limit the use of newer, Apple Silicon-optimized APIs. The focus is on the latest hardware platform for optimal performance.