Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
What This Skill Does
HARPA Grid controls real Chrome browsers remotely through a REST API. It scrapes pages, runs web searches, executes built-in or custom AI commands, and sends AI prompts with full page context. Because it operates inside a live browser session, it reaches authenticated pages and JavaScript-rendered content.
Runs inside a real authenticated Chrome session, so it handles login-gated pages, session cookies, and JavaScript-rendered content that headless scrapers cannot reach.
When to use it
- Extracting product prices from behind-login e-commerce dashboards
- Collecting SERP results for a list of search queries programmatically
- Pulling structured data from JavaScript-heavy pages using CSS or XPath selectors
- Analyzing live webpage content with a custom AI prompt in a single API call
- Monitoring competitor pages for new content or price changes
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: Extract all product titles and prices from my store's admin dashboard
- 1Authenticates the POST request to api.harpa.ai using HARPA_API_KEY
- 2Sends a scrape action targeting the dashboard URL with grab selectors for .product-title and .product-price CSS classes
- 3HARPA browser node opens the page using the existing authenticated browser session
- 4Returns a JSON response with two labeled arrays containing the extracted titles and prices
JSON object with labeled arrays of product titles and corresponding prices from the authenticated dashboard
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.