Claude Skill MartBrowse skillsWhat is a Skill?
 
← Back to list

agent-browser Browser Automation

Lets Claude drive a real browser to navigate sites, fill forms, capture screenshots and extract page data.

AutomationIntermediate★ 67,531⑂ 5,507AI score 9/10Last updated: Aug 9, 2026

What it does

This skill wraps the agent-browser CLI so Claude Code can control a real Chromium browser using the loop open → snapshot → act on refs (@e1).

  • Navigation, history, reload, tabs, windows and iframe switching
  • Accessibility-tree snapshots that return interactive elements with stable refs
  • Full interaction set: click, fill, type, check, select, drag, upload, keyboard/mouse control
  • Screenshots (full page, annotated with numbered labels), PDF export, video recording
  • Cookie/localStorage/sessionStorage access plus saved auth state and profiles
  • Network interception and mocking, console/error inspection, tracing and profiling
  • Diffing snapshots, screenshots or two URLs for visual regression checks
  • Device emulation, iOS Simulator and cloud browser providers (Browserbase, Kernel, Browser Use)

Who it's for

  • Developers who want E2E web tests driven by plain-language instructions
  • Anyone automating repetitive tasks behind a login
  • Frontend engineers comparing UI before and after a change
  • Researchers scraping structured data from public pages

Examples

  1. Login flow test: "Open example.com/login, sign in with the test account and confirm the dashboard loads" → open → snapshot -i → fill @e1/@e2 → click @e3 → wait --url "**/dashboard".
  2. Reuse authentication: log in once, run state save auth.json, then start later sessions with state load auth.json to skip the login entirely.
  3. Visual regression: diff url https://old.example.com https://new.example.com --screenshot to produce a pixel diff between staging and production.

· · · Install guide · · ·

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the packages/omo-opencode/src/features/builtin-skills/agent-browser folder from the GitHub repo code-yeongyu/oh-my-openagent into my ~/.claude/skills/agent-browser/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/code-yeongyu/oh-my-openagent.git /tmp/omo && mkdir -p ~/.claude/skills && cp -r /tmp/omo/packages/omo-opencode/src/features/builtin-skills/agent-browser ~/.claude/skills/ && bun add -g agent-browser && agent-browser install

⚠ This is a third-party skill. Check the source repository before installing.

  1. Check that Bun is available: bun --version (install with curl -fsSL https://bun.sh/install | bash if missing).
  2. Install the CLI globally: bun add -g agent-browser.
  3. Download the browser engine: agent-browser install (use --with-deps on Linux).
  4. Clone the skill repository: git clone https://github.com/code-yeongyu/oh-my-openagent.git /tmp/omo.
  5. Copy the skill folder: mkdir -p ~/.claude/skills && cp -r /tmp/omo/packages/omo-opencode/src/features/builtin-skills/agent-browser ~/.claude/skills/.
  6. Restart Claude Code and try: "Open example.com in the browser and take a screenshot."
  7. Optional: store repeated flags in ~/.agent-browser/config.json, e.g. {"headed": true}, and keep saved state files out of version control.
View source on GitHub ↗License: NOASSERTION