harpa-ai_skill

This skill automates browser tasks and data extraction from live websites via HARPA Grid, enabling scalable scraping, SERP searches, and AI prompts.
  • Python

2.6k

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill harpa-ai

  • _meta.json271 B
  • SKILL.md6.2 KB

Overview

This skill automates real web browsers via the HARPA AI Grid REST API to scrape pages, run searches, execute built-in or custom AI commands, and send AI prompts with full page context. It requires the HARPA AI Chrome extension, at least one active browser node, and a HARPA API key provided as HARPA_API_KEY. The skill exposes a single POST endpoint that orchestrates remote browser actions and can return results synchronously or via webhook.

How this skill works

You call POST https://api.harpa.ai/api/v1/grid with Authorization: Bearer $HARPA_API_KEY and a JSON body specifying action: scrape, serp, command, or prompt. Scrape supports full-page markdown output or targeted selectors (CSS/XPath/text) with flexible extraction options. Command and prompt actions run AI automations or custom prompts using page context ({{page}}) and can target specific nodes or broadcast to many. Results can be returned immediately or posted to a resultsWebhook for asynchronous delivery when nodes are offline or tasks are long-running.

When to use it

  • Extract structured data from public pages or sites where your browser session has access (logged-in content).
  • Run targeted web searches (SERP) and fetch search results programmatically.
  • Execute built-in HARPA automations or custom multi-step commands on live pages.
  • Send AI prompts that include full page context to summarize, extract contacts, or analyze content.
  • Automate repetitive browsing tasks across specific nodes or across all available nodes.

Best practices

  • Install the HARPA Chrome extension and configure at least one active node before calling the API.
  • Use the grab array to collect structured fields in a single request instead of repeated scrapes.
  • Increase timeout for long AI commands or use resultsWebhook to handle asynchronous results.
  • Specify node IDs when you need a particular browser session (cookies, auth state).
  • Use selectorType and take options to precisely capture innerText, href, attributes, or HTML as needed.

Example use cases

  • Scrape product titles and prices from an e-commerce category page using multiple CSS selectors.
  • Run a site-limited SERP query to discover recent articles mentioning a brand.
  • Invoke a custom HARPA command to extract all headings and return word counts from an article.
  • Send an AI prompt that includes {{page}} to extract contact info from a company homepage.
  • Dispatch the same scrape across multiple nodes to parallelize data collection or compare rendered content.

FAQ

Yes. You must install the HARPA AI Chrome extension, configure at least one active node, and use the HARPA_API_KEY from the extension's AUTOMATE tab.

How do I handle long-running tasks or offline nodes?

Set a larger timeout (up to 300000 ms) or provide a resultsWebhook to receive results asynchronously; actions can stay alive up to 30 days.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational