AIAny
AI Agent2025
Icon for item

VibeTunnel

Forwards local terminal sessions to any web browser, so you can watch and steer long-running CLI processes — including AI coding agents like Claude Code — from a phone or another machine. A macOS menu-bar app proxies PTY output over WebSocket.

Introduction

AI coding agents quietly changed what people want from a terminal. When Claude Code or Codex grinds away unattended for ten minutes, the real constraint isn't typing — it's being tethered to the Mac that's running them. VibeTunnel inverts that: the terminal follows you, streaming a live local shell into any browser so you can approve a prompt or kill a runaway agent from your phone.

What Sets It Apart
  • It forwards a real PTY, not a log tail. Output and input are proxied over WebSocket, so the browser gets a fully interactive session — scrollback, control keys, resizing — rather than a read-only feed.
  • The vt wrapper makes adoption nearly free: prefix any command (vt claude, vt npm run dev) and that session becomes reachable, with no per-tool integration.
  • Git follow mode keeps a main checkout in sync with whichever worktree branch you switch to, which matters when an agent is hopping between branches while you watch remotely.
  • Sessions record to asciinema format, so a long agent run can be replayed and shared instead of scrolling a dead buffer.
Where It Fits

It sits between raw SSH (powerful but clumsy on a phone) and screen-sharing (heavy, not terminal-aware). The menu-bar app on macOS is the polished path; a headless npm package covers Linux servers and CI-style boxes.

Great Fit / Look Elsewhere

Great fit if you run autonomous coding agents or long builds on a Mac and want to supervise them from a browser or phone, or expose one machine's terminal across your network. Look elsewhere if you need Windows (unsupported), an Intel Mac native app (Apple Silicon only), or a hardened multi-user remote-shell service — the iOS app is still work-in-progress and the security model assumes you control the network or tunnel.

Information

  • Websitegithub.com
  • OrganizationsAmantus Machina
  • Authorsamantus-ai, Mario Zechner, Armin Ronacher, Peter Steinberger, Helmut Januschka, Manuel Maly
  • Published date2025/06/15

More Items

Hugging Face
AI Model2026

Open-weights LLM fine-tuned for phone-based voice agents that prioritizes low latency and reliable tool/function calling. Based on NVIDIA Nemotron 3 Nano (30B total, 3.5B active), supports very long contexts (262,144 tokens) and recommends temperature=0 with thinking disabled for deployment.

GitHub
AI Infra2025

Measures generative AI inference performance with token-level metrics (TTFT, inter-token latency), latency, and throughput under realistic traffic patterns. Provides a multiprocess engine, real-time TUI dashboard, extensible plugins, and integrations for telemetry and result uploads, aimed at inference benchmarking and capacity planning.

GitHub
AI Agent2026

A curated collection of production-ready Agent Skills that turn tasks—presentation production, image generation, local KB retrieval, article assembly, and web-design—into agent-loadable skill folders. Uses a SKILL.md contract, supports multiple agent runtimes (Claude Code, Cursor, Codex), and offers modular install paths with pinned release zips.