skillfed

agent-browser

agent-browser enables autonomous web automation by executing browser tasks through an intelligent agent runtime. Streamline repetitive workflows like form submission, element interaction, and page navigation without manual intervention. Built for developers who need reliable, scriptable control over web-based processes.

agent-browser is a browser automation tool that enables autonomous web automation by executing browser tasks through an intelligent agent runtime. It streamlines repetitive workflows like form submission, element interaction, and page navigation without manual intervention. The skill is built for developers who need reliable, scriptable control over web-based processes, supporting tasks such as clicking buttons, filling forms programmatically, and automating website interactions.

AI-generated summary based on this skill's SKILL.md

722 167 MIT updated by OtterMind

Install

OtterMind/youclaw/agent-browser · repository language: TypeScript

CLI (skillfed)coming soon
git clone https://github.com/OtterMind/youclaw
cp -r youclaw/skills/agent-browser ~/.claude/skills/agent-browser

Frequently asked questions

AI-generated answers based on this skill's SKILL.md and metadata

What is agent-browser and what can it automate?

agent-browser is a browser automation tool that enables autonomous web automation by executing browser tasks through an intelligent agent runtime. It streamlines repetitive workflows like form submission, element interaction, and page navigation without manual intervention. The skill is built for developers who need reliable, scriptable control over web-based processes, supporting tasks such as clicking buttons, filling forms programmatically, and automating website interactions.

Can agent-browser click buttons and fill forms programmatically?

Yes, agent-browser is designed specifically for automating repetitive browser tasks like form filling, clicking, and navigation. You can programmatically interact with web elements, submit forms automatically, and control browser behavior through its command-line interface. This makes agent-browser ideal for batch browser commands and DOM element interaction without requiring manual intervention.

How does agent-browser help with web scraping and data extraction?

agent-browser enables you to extract data and take screenshots from websites programmatically. You can open URLs, capture snapshots of web pages, and retrieve information from the DOM automatically. This functionality makes agent-browser a powerful tool for web scraping tasks and automated data collection workflows.

Can agent-browser test web applications automatically?

Yes, agent-browser supports testing web applications and verifying page state changes. As a web app testing tool and automated web testing framework, it allows you to validate application behavior, confirm element states, and ensure web pages respond correctly to automated interactions. This makes it suitable for continuous integration and automated quality assurance workflows.

Does agent-browser support persistent login sessions?

agent-browser enables you to login to websites programmatically and persist authentication state across interactions. This capability allows you to reuse login sessions and maintain authenticated state throughout your automation workflows, eliminating the need to re-authenticate for each task.

What license does agent-browser use?

agent-browser is released under the MIT license, which permits free use, modification, and distribution of the software with minimal restrictions.

SKILL.md

rendered from the published skill — quoted content, verbatim

Browser Automation with agent-browser

Performance Rules (CRITICAL)

  • ALWAYS chain commands with && when you don't need intermediate output. Each separate tool call costs seconds of round-trip latency.
  • ALWAYS combine open + wait + snapshot into one call: agent-browser open <url> && agent-browser wait --load load && agent-browser snapshot -i
  • ALWAYS batch multiple interactions (fill, click, select) into one && chain when refs are already known.
  • Use --load load (DOM load event) by default. Only use networkidle when you specifically need all XHR/fetch to complete (e.g., waiting for API-driven content).
  • Do NOT snapshot after every interaction — only re-snapshot when you need to discover new element refs (after navigation or major DOM changes).

Core Workflow

Every browser automation

(truncated - see the full file via the links below)

Read as markdown · JSON record · Browse the source repository

File tree — 1 file
skills/agent-browser/SKILL.md

Related skills

Tags

headless-browser form-automation web-scraping cli-tool element-interaction session-management batch-commands screenshot-capture login-automation dom-manipulation