l0ng-ai / tty7

A GPU-rendered terminal where shells outlive the window and AI agents can orchestrate other agents through four simple CLI commands.

活跃维护 Apache 2.0 Rust Tracked
1.3k 117 3 小时前
CI
alacritty cross-platform daemon developer-tools gpu gpui linux macos pty rust shell terminal terminal-emulator windows ai-agents claude-code coding-agent ssh terminal-multiplexer tmux

星标趋势

数据积累中,暂无足够数据生成趋势图

AI 分析

项目摘要

tty7 is a GPU-rendered terminal emulator built in Rust that decouples shell sessions from the application window, allowing shells to survive quitting the app. It features built-in orchestration for 26 AI coding agents (Claude Code, Codex, Gemini, etc.) with a unified sidebar showing agent status, and enables agents to drive other agents through simple CLI commands without requiring tmux or additional configuration.

为什么值得关注

It solves the persistent session problem natively without tmux while adding first-class AI agent orchestration—agents can spawn, monitor, and control other agents via four simple CLI commands, creating a programmable multi-agent workflow that works with or without the GUI open.

优势

  • Persistent shells that survive window close and reboot without tmux
  • Native orchestration for 26 AI coding agents with status tracking
  • Agent-to-agent automation via simple CLI commands (split, send, wait)
  • High-performance GPU rendering using Zed's gpui framework
  • Cross-platform (macOS, Windows, Linux) with pure Rust implementation
  • Very active development with 33 releases in 6 months

局限性

  • No automated tests despite active development
  • Very new project (created July 2026) with limited production track record
  • No Docker support or official examples
  • Documentation beyond README appears limited

使用场景

  • Managing multiple concurrent AI coding agents across different repositories
  • Remote development with persistent sessions that survive disconnections
  • Building automated agent-to-agent workflows for CI/CD or code review
  • Replacing tmux for users who want session persistence without configuration overhead
目标用户: Software developers and AI engineers who use multiple coding agents (Claude Code, Codex, Gemini CLI) and need better orchestration, visibility, and session persistence than traditional terminals provide.
学习曲线: 中
分析模型:LongCat-2.0 | 分析时间:25 天前