A sophisticated AI-driven browser automation skill that leverages Amazon Nova Act to execute complex web tasks via natural language commands.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install nova-act
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 nova-act 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 Nova Act Browser Automation?
The Nova Act skill enables AI agents to interact with web browsers with human-like precision. Built on the Amazon Nova Act framework, this tool allows developers to automate intricate workflows such as flight searches, data extraction, and multi-step form navigation using simple natural language. As a key component of the Openclaw Skills ecosystem, it provides a bridge between high-level reasoning and low-level browser interactions.
This skill is designed with a focus on reliability and developer experience. By integrating this into your Openclaw Skills library, you can write Python scripts that not only navigate the web but also extract structured data using Pydantic schemas. It handles the complexities of page stability and element selection, allowing you to focus on the logic of your automation tasks.
Nova Act Browser Automation Use Cases
- Searching and comparing travel itineraries across multiple booking platforms.
- Scraping structured data from dynamic websites into JSON or Pydantic models.
- Automating the navigation of complex web applications to verify UI states.
- Pre-filling contact or registration forms for manual verification.
- Performing research tasks that require interacting with paginated search results.
How Nova Act Browser Automation Works
- The skill initializes a connection to Amazon Nova Act using the provided API key.
- It launches a managed Chrome browser instance through the Playwright framework.
- The AI agent processes natural language prompts to generate browser actions via the
nova.act()method. - For data-centric tasks,
nova.act_get()retrieves structured information based on a user-defined schema. - Built-in safety logic scans for material-impact keywords to prevent unauthorized real-world actions.
- The session concludes by generating detailed trace files and screenshots for auditability.
Nova Act Browser Automation Setup
To integrate this into your collection of Openclaw Skills, you must have the uv package manager installed and a valid API key from Amazon Nova.
# Install the required package manager
brew install uv
# Configure your environment variable
export NOVA_ACT_API_KEY='your_api_key_here'
Alternatively, you can store your credentials in the global configuration at ~/.openclaw/openclaw.json to be used by all compatible Openclaw Skills.
Nova Act Browser Automation Data Schema & Taxonomy
The skill organizes its output and session data to ensure transparency and ease of use:
| Component | Description | Format |
|---|---|---|
| Trace Files | Full session recordings and screenshots | .png / .trace |
| Extracted Data | Structured output from web pages | JSON / Pydantic |
| Safety Logs | Documentation of any triggered safety stops | Console Output |
| Environment | API configuration and session metadata | JSON |
All trace files are stored in the current working directory, allowing for immediate review of the AI agent's decisions.
Nova Act Browser Automation Advanced Features
- Structured data extraction support using Python types and Pydantic models for high-accuracy scraping.
- Material-impact guardrails that automatically halt the agent before actions like purchasing, posting, or deleting.
- PEP 723 script support allowing for portable automation scripts with inline dependency management.
- Headless and headed mode support for both background processing and visual debugging.
- Direct access to the underlying Playwright page object for advanced DOM manipulation and custom workflow extensions.
SKILL.md
Loading
Related Openclaw Skills
Computer Vision Expert
An advanced vision systems architect specialized in real-time detection, promptable segmentation, and 3D spatial reasoning using 2026 state-of-the-art models.
Sinkron
Sinkron enables AI agents to possess permanent email addresses and manage inboxes programmatically via a CLI and Python SDK.
Cloudflare R2 CLI
A minimal Python-based CLI tool for interacting with Cloudflare R2 storage using secure S3-compatible APIs and AWS Signature V4.
MailTap Temporary Email Service
A developer-friendly API for generating disposable 30-minute email addresses and retrieving messages without authentication.
deAPI AI Media Suite
A cost-effective, decentralized AI toolkit for processing transcription, images, video, and audio through a unified API.
Agentic Money
A decentralized protocol skill enabling AI agents to discover, hire, and receive payments for services via Ethereum-based escrow.
METADATA
Featured*
