browserless-agent

What This Skill Does

The browserless-agent is a professional-grade web automation skill for the OpenClaw ecosystem. It leverages headless browser technology to perform complex tasks that mimic real human interaction with the web. From simple navigation and text scraping to advanced operations like file uploads, multi-tab management, and network request interception, this skill transforms the AI agent into a digital worker. It supports high-fidelity visual capturing through screenshots and PDFs, allowing users to archive content or audit visual layouts automatically. By providing granular control over browser state, geolocation, and device emulation, it ensures that automation tasks are robust and bypass common bot-detection barriers.

Installation

To install this skill, run the following command in your terminal: clawhub install openclaw/skills/skills/raff-lima/browserless-agent. After installation, navigate to your OpenClaw settings under the Skills tab. You must provide the BROWSERLESS_URL for your headless browser instance. If you are using a cloud-hosted provider like Browserless.io, you should also include your BROWSERLESS_TOKEN in the environment variables configuration. The system will automatically append the necessary connection paths and authentication tokens to ensure a secure and stable connection to your browser instance.

Use Cases

This skill is ideal for e-commerce price monitoring, automated quality assurance testing, lead generation through directory scraping, and generating automated PDF reports from dynamic dashboards. It is also highly effective for social media management tasks, such as filling out web-based forms or navigating complex multi-step login portals that require cookie and session management.

Example Prompts

1. "Open https://tech-crunch.com, wait for the page to load, and scrape all article titles and their corresponding URLs into a list."
2. "Navigate to the project management dashboard, log in with my saved credentials, capture a full-page screenshot, and save it as a PDF report."
3. "Go to the search page, type 'OpenClaw documentation' into the search bar, submit the form, and return the text content of the first three results."

Tips & Limitations

Always ensure your Browserless instance has enough resources to handle complex DOM structures. Avoid excessive concurrency if using a free-tier hosting service. When scraping, define specific selectors rather than generic tags to improve accuracy. Note that this skill requires external network connectivity to your specified Browserless URL, and it is subject to the anti-scraping policies of the target websites you interact with.