AgentIndex icon
AgentIndex
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Submit Tool
ToolsMCP ServersHOTCategoriesTrendingNewBest ListsCompareGuides
Home/
Browser Automation/
terminal-browser
terminal-browser logo

terminal-browser

Active·★ 4.7k·MIT·Updated 2026-11-21

### Architecture & Positioning terminal-browser, developed by Zenbu Labs, bridges the gap between text-based terminal interfaces and graphical web browsers. Utilizing the Kitty Graphics Protocol, it renders a full-fledged, real Chromium browser directly inside modern GPU-accelerated terminal emulators. As terminal-first AI coding agents (such as Claude Code and Aider) become mainstream, context switching between CLI terminals and desktop GUI browsers introduces friction; terminal-browser solves this by rasterizing Chromium frames and streaming them via CDP into terminal pane layouts alongside the active coding assistant. ### Core Capabilities & Differentiation Unlike traditional terminal browsers like Lynx or w3m that only parse raw text, terminal-browser runs a full Chromium engine supporting complex JavaScript, CSS3, dynamic DOM rendering, and interactive developer tools. Its key differentiator is two-fold: first, native support for tabs, split-pane previews, and remote SSH forwarding; second, a built-in agent-browser-compatible Action CLI that allows AI agents to inspect open tabs, issue click and typing events, scroll viewports, and capture screenshots directly through standard terminal commands. ### Deployment & Best Practices To run terminal-browser smoothly, ensure you use a terminal that fully supports the Kitty graphics protocol (such as Kitty, Ghostty, or configured WezTerm). For remote development over SSH, ensure high-bandwidth network connectivity and proper socket forwarding to avoid visual lag. When integrating with automated coding agents, configure the Action CLI as an executable tool with built-in DOM-ready wait states, guaranteeing that heavy React/Vue SPA hydration finishes before agent click dispatches.

This write-up is based on the project's README and public repository data. We have not hand-tested the tool.

#browser#claude-code#claude-code-plugin#claude-skills#cli#codex#dev-tooling#herdr-plugin
$ Install
$ brew install terminal-browser
↗ Visit site★ GitHub
01

Features

01Renders browser pixels through the kitty graphics protocol.
02Provides tabs, split panes and developer tools.
03Action commands let agents interact with open browsers.
02

Why choose it

+Renders full-featured Chromium with rich JavaScript, CSS3, and DOM support directly inside graphics-capable terminal windows.
+Eliminates window-switching friction by enabling side-by-side split previews alongside active AI coding agents.
+Built-in Action CLI compatible with agent-browser allows programmatic click, typing, scroll, and inspection from the shell.
+Supports remote development workflows over SSH tunnels without requiring X11 forwarding or heavy VNC desktop setups.
03

Trade-offs

!Strictly requires terminal emulators supporting the Kitty graphics protocol, rendering it incompatible with standard macOS Terminal or basic PuTTY.
!Higher CPU, memory, and GPU usage compared to headless browsers due to continuous frame rasterization and terminal protocol encoding.
!High network latency on remote SSH links can lead to frame drops and sluggish mouse/keyboard responsiveness.
!Complex multi-monitor scaling configurations on Linux Wayland or macOS may require manual font and cell dimension calibration.
04

Maintenance check

ActiveChecked: 2026-11-13
Last commit
2026-11-13
Commits, last 90 days
100+
Releases, last 90 days
30
Latest release
pixel-v0.2.21 (2026-11-13)
Open issues
135
Recent issues with at least one comment
2/24
Commits per week, last 12 weeks
Commits per week, last 12 weeks: 9, 7, 1, 34, 1, 18, 3, 5, 57, 13, 6, 13

Oldest week on the left; the latest four weeks are orange. Source: GitHub commit activity.

Status rule: archived on GitHub = Archived; last commit within 90 days = Active; 91 to 365 days = Slowing; older = Stale. All figures come from the GitHub API on the date shown.

Most discussed open issues

  • #89 Herdr - Terminal-Browser not rendering8 comments · 2026-08-30
  • #207 terminal-browser 中文乱码5 comments · 2026-10-26
  • #97 herdr graphics client is dark on herdr 0.8.2, and inline set cannot replace the file transport4 comments · 2026-09-05
  • #16 Pages appear zoomed ~2x when Ghostty is on a secondary 1x display (hostDisplayScale uses primary only)4 comments · 2026-08-04
  • #126 agent-browser daemons survive shutdown and are never reaped3 comments · 2026-09-27

Issue titles are quoted from GitHub. Open the link to check whether a problem still applies to the version you use.

05

Compatibility

macos
Requires a supported graphics-capable terminal
Verified via docs
linux
Requires a supported graphics-capable terminal
Verified via docs
windows
WSL required; terminal graphics support is limited
Verified via docs
06

Quick start

1
$ brew install terminal-browser
07

Use cases

↳Preview a website beside a coding agent.
↳Inspect a remote development site over SSH.
↳Operate a browser through compatible agent commands.
08

Alternatives

⚡Top VS:
VSdarwin-skill★7.0k→VSanswer-me-with-html★6.6k→VSmodlens★5.0k→VSobsidian-wiki★3.8k→
marketingskills logo
marketingskills★ 61.9k
Marketing skills for Claude Code and AI agents. CRO, copywriting, SEO, analytics, and growth engineering.
cmux logo
cmux★ 28.8k
Open source Ghostty-based macOS terminal with vertical tabs and notifications for AI coding agents. Built for multitasking, organization, and programmability.
video-shotcraft logo
video-shotcraft★ 22.0k
AI video skill for Claude Code & Codex — cinematic, editable product promos with Remotion: 124 shot recipe cards, 150 motion previews, beat sync, sound design, Motion Workbench & JianYing (CapCut CN) export.
darwin-skill logo
darwin-skill★ 7.0k
达尔文.skill —— 一个让你的Skill无限进化的系统:评估→改进→测试→保留或回滚 | Autoresearch-inspired autonomous skill optimization for Claude Code. Evaluate, improve, test, keep or revert.
VSCompare→
answer-me-with-html logo
answer-me-with-html★ 6.6k
Answer me with HTML — an agent skill that answers hard questions with a one-page HTML you can actually read. 让 AI Agent 用一页 HTML 回答复杂问题。
VSCompare→
modlens logo
modlens★ 5.0k
The first vision plugin for DeepSeek Harness, and the vision bridge for every text-only coding agent. Paste an image, get structured JSON evidence (OCR, layout, semantics). | 全网最强 DeepSeek Harness 外挂视觉插件,为 DeepSeek、GLM 等纯文本模型外挂视觉能力,粘贴图片即得结构化 JSON 证据(OCR、版面、语义)。
VSCompare→
obsidian-wiki logo
obsidian-wiki★ 3.8k
Framework for AI agents to build and maintain a digital brain through Obsidian wiki | Memory System for Agents
VSCompare→
huashu-motion logo
huashu-motion★ 2.9k
让你的agent一键直出商业级产品宣传片:会作曲,能指着画面改。Commercial-grade launch films by your coding agent — point at the frame to change it.
VSCompare→
See all alternatives →

Comments

Log in to leave a comment

No comments yet. Be the first!

On this page
01Features02Why choose it03Trade-offs04Maintenance check05Compatibility06Quick start07Use cases08Alternatives
zenbu-labs /
terminal-browser
↗

Native Chromium browser rendered directly inside graphics-capable terminals via Kitty protocol, featuring split-pane live preview and agent action CLI

4.7kMIT
Last commit1d ago
StatusActive
CategoryBrowser Automation
Links
Documentation↗Discussion↗Issues↗Releases↗
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
© 2026 AgentIndex.app|Built by a 10-year iOS Developer.
Best by use caseAboutContactAdvertisePrivacy PolicyTerms of UseQYSGitHubBuy me a coffee ☕

Browse by Category

MCP ServersCode AssistantWorkflow AutomationRAG / Knowledge BaseMulti-AgentBrowser AutomationLLM InfraDev ToolingObservability

Not affiliated with Anthropic, OpenAI or Microsoft.