agent-browser Automation

Teaches Claude to drive the `agent-browser` CLI to open sites, fill forms, click, screenshot, and extract data reliably.

AutomationIntermediate★ 68,873⑂ 6,858AI score 9/10Last updated: Nov 9, 2026

What it does

Encodes the core automation loop: open → snapshot -i to get element refs (@e1, @e2) → click/fill/select → re-snapshot after any DOM or page change.
Provides a compact command reference for navigation, interaction, waits (wait --load networkidle, wait --url), text/URL/title extraction, screenshots and PDF export.
Ships battle-tested patterns: form submission, login with persisted state (state save/load), JSON data extraction, parallel named sessions, headed debugging with video recording, proxies, local file:// access, and iOS Simulator (mobile Safari) testing.
Explains ref invalidation rules and offers semantic locators (find text/label/role/testid) as a fallback.
Includes three runnable shell templates: form automation, authenticated session, capture workflow.

Who it's for

Developers and QA engineers automating repetitive web app checks.
Anyone scraping or monitoring pages that sit behind a login.
Teams generating documentation screenshots, PDFs, or demo recordings.
Prerequisites: comfort with a terminal and the agent-browser CLI installed (iOS flows need macOS + Xcode + Appium).

Example uses

"Fill this signup form with test data, submit it, and screenshot the result" → open → snapshot -i → fill/select/check → click → wait --load networkidle → screenshot.
"Log into the admin app once and reuse the session next time" → login, then state save auth.json; later state load auth.json and go straight to the dashboard.
"Compare pricing tables on two competitor sites at once" → run --session site1 and --session site2 in parallel and pull snapshot -i --json from each.

· · · Install guide · · ·

Try it now, no install

Paste this into Claude to use the skill without installing anything.

Read the instructions in this file and follow them to help me:
https://raw.githubusercontent.com/shanraisshan/claude-code-best-practice/HEAD/.claude/skills/agent-browser/SKILL.md

What I want: (describe your task here)

If Claude can't open the link, open it yourself and paste the contents instead.

↓ If it works for you, download the ZIP below and install it. Then it runs on its own — no pasting each time.

Install in the Claude app (no terminal)

Download the ZIP with the button below.
In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
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 AI Skill Mart.
Copy the .claude/skills/agent-browser folder from the GitHub repo shanraisshan/claude-code-best-practice into my ~/.claude/skills/shanraisshan-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/shanraisshan/claude-code-best-practice.git /tmp/ccbp && mkdir -p ~/.claude/skills && cp -r /tmp/ccbp/.claude/skills/agent-browser ~/.claude/skills/

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

Open your terminal.
Clone the repo: git clone https://github.com/shanraisshan/claude-code-best-practice.git /tmp/ccbp
Copy the skill into Claude Code: mkdir -p ~/.claude/skills && cp -r /tmp/ccbp/.claude/skills/agent-browser ~/.claude/skills/
Install the agent-browser CLI itself (follow the tool's official instructions) and verify with agent-browser --help.
Optional: for iOS Simulator testing on macOS, install Xcode and run npm install -g appium && appium driver install xcuitest.
Restart Claude Code, then try a prompt like "open this URL and take a full-page screenshot" to trigger the skill.
To run the bundled scripts directly: chmod +x ~/.claude/skills/agent-browser/templates/*.sh.

View source on GitHub ↗License: MIT