agent-browser

Automate browser navigation, form filling, screenshots, and data extraction via CLI commands.

Updated Apr 13, 2026
One-click install
npx skills add https://github.com/tsmes/conaro --skill agent-browser-tsmes
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-browser
Source: https://github.com/tsmes/conaro/tree/main/.claude/skills/agent-browser
Command: npx skills add https://github.com/tsmes/conaro --skill agent-browser-tsmes

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Manually interacting with websites for testing, scraping, or repetitive form submission is slow and error-prone, and raw DOM parsing consumes excessive context for AI agents. This Skill provides a CLI-driven browser automation workflow where compact element refs replace verbose HTML parsing. ## Core Features & Use Cases - Snapshot-Based Interaction: Capture interactive elements as compact refs (@e1, @e2) and click, fill, select, or check them directly, reducing token usage compared to full DOM parsing. - Session & State Management: Run isolated named sessions in parallel, save and restore authentication state, and persist cookies across runs. - Capture & Verification: Take annotated screenshots, record video, save pages as PDF, and diff snapshots or screenshots to verify changes. - Use Case: An AI agent needs to log into a web app, fill out a multi-field form, and verify the submission succeeded. It opens the login page, snapshots to get refs, fills credentials, saves the auth state, then reuses that state for subsequent automated runs. ## Quick Start Use the agent-browser skill to open https://example.com/login, fill in the email and password fields, click submit, and take a screenshot of the result.

Frequently Asked Questions about agent-browser

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

FAQPage Schema
How do I automate filling out a web form from the command line?▼

Open the page with agent-browser open, run snapshot -i to get element refs like @e1, then use fill, select, and check commands on those refs before clicking the submit button. Re-snapshot after submission to verify the result.

How do I save and reuse login sessions in browser automation?▼

After completing a login flow, run agent-browser state save auth.json to persist cookies and storage. Later runs restore it with state load auth.json, letting you skip the login flow and navigate directly to protected pages.

Why do element refs stop working after clicking a button?▼

Refs are invalidated whenever the page changes, including navigation, form submissions, and dynamic content updates. Always run snapshot -i again after any action that modifies the DOM to get fresh refs before interacting further.

Can I run multiple browser automation sessions in parallel?▼

Yes, use the --session flag with distinct names to create isolated browser contexts with separate cookies, storage, and tabs. This supports concurrent scraping or A/B testing, and sessions should be closed individually when finished.

Does agent-browser support mobile or iOS testing?▼

Yes, it supports iOS Simulator via the -p ios flag with a device name, enabling tap, swipe, and screenshot commands on Mobile Safari. This requires macOS with Xcode and Appium with the xcuitest driver installed.

How do I avoid shell quoting errors when running JavaScript in the browser?▼

Use eval --stdin with a heredoc or eval -b with base64-encoded scripts instead of inline quoted expressions. These flags bypass shell interpretation, preventing corruption from nested quotes, backticks, or special characters.