A high-performance browser automation library providing standardized web interaction, navigation, and data capture capabilities for AI agents.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install browser-automation-core
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install browser-automation-core using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Browser Automation Core?
The Browser Automation Core serves as the foundational interaction layer for web-based AI agents. It provides a robust, reusable suite of tools for navigating complex websites, interacting with DOM elements, and capturing data, allowing developers to build sophisticated agents without rebuilding core browser logic. This skill is a central component of the Openclaw Skills ecosystem, specifically designed to power specialized agents like Facet for CAD learning and Ace for competition entry.
By leveraging the Chrome DevTools Protocol (CDP), this skill offers deeper control over browser behavior than traditional automation frameworks. It manages the entire lifecycle of a web session, from initial authentication and cookie management to complex multi-tab workflows and error recovery, ensuring that your Openclaw Skills operate reliably in dynamic web environments.
Browser Automation Core Use Cases
- Automating complex form filling and data submission for competition entries.
- Navigating multi-step educational tutorials on platforms like Onshape for CAD learning.
- Performing high-fidelity visual regression testing and screenshot capture for web audits.
- Managing persistent user sessions and authentication states across multiple agent tasks.
- Orchestrating parallel browser interactions across several tabs and windows.
How Browser Automation Core Works
- The skill establishes a connection to the integrated browser via a WebSocket using the Chrome DevTools Protocol (CDP) on port 18800.
- It loads environment-specific configurations such as headless mode, viewport dimensions, and custom user agents.
- High-level agent commands like navigation or form submission are translated into precise CDP methods for direct browser manipulation.
- Intelligent wait conditions monitor the page state to ensure elements are visible and interactable before actions are performed.
- The skill implements an error recovery layer that handles network retries and utilizes fallback selectors if primary elements are not found.
- All session data, screenshots, and extracted HTML are organized and saved to the designated Openclaw Skills workspace for further processing.
Browser Automation Core Setup
To integrate this core functionality into your Openclaw Skills, you can install it via ClawHub:
npx clawhub@latest install browser-automation-core
For local development, copy the skill directory into your workspace:
cp -r /path/to/skill /root/.openclaw/workspace/skills/
Ensure your environment is configured with the necessary CDP parameters:
export BROWSER_HEADLESS="true"
export CDP_URL="http://localhost:18800/json"
export BROWSER_TIMEOUT="30"
Browser Automation Core Data Schema & Taxonomy
The Browser Automation Core organizes its operations into a clean directory structure and configuration schema within the Openclaw Skills framework:
| Directory | Functionality |
|---|---|
/navigation |
Handles URL loading, timeout logic, and page state waits. |
/interaction |
Manages click, type, hover, and keyboard event sequences. |
/capture |
Stores screenshots, HTML source, and extracted text metadata. |
/forms |
Contains logic for form detection, field mapping, and validation. |
/sessions |
Manages cookie storage and authentication persistence. |
Configuration is handled via a JSON5 schema that defines CDP endpoints, headless preferences, and directory paths for stored media.
Browser Automation Core Advanced Features
- Direct Chrome DevTools Protocol (CDP) integration for granular browser control beyond standard DOM interactions.
- Performance-optimized command batching to reduce latency during high-frequency browser operations.
- Sophisticated error recovery patterns including multi-selector fallback strategies and automatic session restoration.
- Modular architecture designed for agent-specific extensions, supporting complex domain-specific logic like Onshape CAD automation.
- Comprehensive session isolation to ensure security and privacy when running multiple Openclaw Skills simultaneously.
SKILL.md
Loading
Related Openclaw Skills
Skill Creator
A comprehensive framework for creating, refining, and packaging modular AI agent skills to extend coding assistant capabilities.
X Engagement
A growth-oriented skill for AI agents to execute founder-style engagement, audience-building loops, and authentic community participation on X.
text2qa
A powerful extraction skill that converts messy conversational history or raw text into structured Q&A pairs and documented user preferences.
Video Short Creator
A two-phase AI-powered workflow for generating narrated short-form videos with human review checkpoints and automated FFmpeg assembly.
Mimic AI Character Generator
Mimic is a powerful character creation skill that analyzes real-world speech patterns and social media data to turn your AI into any specific individual.
WeShop AI Sexy Teens OpenAPI Skill
A specialized AI skill for transforming portrait or full-body photos into high-quality bikini model images and videos via the WeShop OpenAPI.
Featured*
