agent-browser

Automates browser interactions via Chrome DevTools Protocol with accessibility-tree snapshots and element references.

2|1|Updated Jun 9, 2026
One-click install
npx skills add https://github.com/zester4/zilmate --skill agent-browser-zester4
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/zester4/zilmate/tree/main/.agents/skills/agent-browser
Command: npx skills add https://github.com/zester4/zilmate --skill agent-browser-zester4

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires agent-browser.

What problem does it solve? AI agents need a reliable way to interact with websites and desktop apps—navigating pages, filling forms, clicking buttons, and extracting data—without heavy dependencies like Playwright or Puppeteer. ## Core Features & Use Cases - Browser Automation via CDP: Control Chrome/Chromium through the DevTools Protocol with accessibility-tree snapshots and compact @eN element references for reliable interaction. - Specialized Skills: Load dedicated workflows for Electron desktop apps (VS Code, Slack, Discord, Figma), Slack workspace automation, exploratory testing/QA, Vercel Sandbox microVMs, and AWS Bedrock AgentCore cloud browsers. - Sessions & Observability: Supports sessions, an authentication vault, state persistence, video recording, and an observability dashboard on port 4848. - Use Case: Ask your agent to log into a web app, fill out a signup form, take a screenshot of the result, and extract the confirmation data—all through CLI commands. ## Quick Start Install the CLI with npm, run agent-browser install, then ask your agent to load the core skill with agent-browser skills get core before automating any browser task.

Frequently Asked Questions about agent-browser

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate browser tasks with an AI agent?

Install agent-browser globally with npm, run agent-browser install, then load the core workflow with agent-browser skills get core. The CLI controls Chrome/Chromium via CDP using accessibility-tree snapshots and element refs for reliable clicking, typing, and navigation.

What browser automation tool works without Playwright or Puppeteer?

agent-browser is a native Rust CLI that drives Chrome/Chromium directly through the Chrome DevTools Protocol, with no Playwright or Puppeteer dependency. It provides accessibility-tree snapshots with compact @eN element references for interaction.

Can agent-browser automate Electron desktop apps like Slack or VS Code?

Yes, load the electron specialized skill with agent-browser skills get electron to automate Electron apps such as VS Code, Slack, Discord, Figma, Notion, and Spotify. A dedicated slack skill also covers unreads, messages, and conversation search.

Does agent-browser support cloud browsers or sandboxed environments?

Yes, it supports running inside Vercel Sandbox microVMs via the vercel-sandbox skill and AWS Bedrock AgentCore cloud browsers via the agentcore skill. Load them with agent-browser skills get followed by the skill name.

How do I keep agent-browser instructions up to date?

The SKILL.md file is only a discovery stub. Always run agent-browser skills get core before use, since the CLI serves skill content matching the installed version so instructions never go stale.