Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.scrapegraphai.com/llms.txt

Use this file to discover all available pages before exploring further.

​
Overview

The ScrapeGraphAI Claude Code Skill ships with just-scrape, the official CLI for the v2 API. Once installed, agents like Claude Code, Cursor, Copilot, Cline, and Windsurf can scrape websites, extract structured data, search the web, crawl sites, and set up page-change monitors — all from natural language prompts. The skill wires just-scrape into your agent’s skill directory so the agent knows when and how to invoke the CLI.

GitHub Repository

Browse the CLI and skill source

​
Installation

2
The fastest way to install. Requires Node.js or Bun.
3
bunx skills add https://github.com/ScrapeGraphAI/just-scrape
# or
npx skills add https://github.com/ScrapeGraphAI/just-scrape
4
This symlinks skills/just-scrape/SKILL.md into your ~/.claude/skills/ directory automatically.
5
You can also browse the published skill at skills.sh/scrapegraphai/just-scrape/just-scrape.
6
Option 2: Manual install
7
Clone the repository and create the symlink yourself:
8
git clone https://github.com/ScrapeGraphAI/just-scrape.git ~/.claude/skills/just-scrape
ln -sf ~/.claude/skills/just-scrape/skills/just-scrape/SKILL.md ~/.claude/skills/just-scrape.md
9
Option 3: Project-level install
10
Install the skill for a single project only:
11
mkdir -p .claude/skills
git clone https://github.com/ScrapeGraphAI/just-scrape.git .claude/skills/just-scrape
ln -sf .claude/skills/just-scrape/skills/just-scrape/SKILL.md .claude/skills/just-scrape.md

​
Setup

Install the CLI and set your ScrapeGraphAI API key: 
npm install -g just-scrape@latest
export SGAI_API_KEY="sgai-..."
Get your API key from the dashboard. The CLI also accepts the key via a .env file, ~/.scrapegraphai/config.json, or an interactive prompt.

​
Capabilities

The skill maps to the v2 API surface via just-scrape:

extract

Extract structured data from any URL using AI (just-scrape extract)

search

Search the web and extract structured results (just-scrape search)

scrape

Fetch a page in 8 formats: markdown, html, screenshot, branding, links, images, summary, json

markdownify

Convert any webpage into clean markdown (wraps scrape -f markdown)

crawl

Crawl multi-page sites with depth, link, and pattern controls

monitor

Schedule page-change monitors with cron intervals, webhooks, and activity polling
Removed from v1: sitemap, agentic_scraper, generate-schema, validate. There is no direct replacement on v2.

​
Example Prompts

Once the skill is installed, you can use natural language prompts directly in your AI coding agent: 
Scrape https://example.com/products and extract all product names, prices, and ratings

Search the web for "best React UI libraries 2025" and extract the top 5 with pros and cons

Convert https://example.com/docs/getting-started to markdown and save it to docs/

Crawl https://example.com/blog with depth 2 and extract the title and summary from each post

Monitor https://store.example.com/pricing every hour and webhook me when it changes

Create a 30m monitor on https://example.com and poll its activity feed, printing new ticks as they come in

Fetch a full-page screenshot and branding assets for https://example.com
The agent will automatically select the right just-scrape command, handle authentication, poll for async results (crawls), and return structured data.

​
Supported Agents

This skill works with any AI coding agent that supports the skills protocol:
  • Claude Code — Anthropic’s CLI agent
  • Cursor — AI-powered code editor
  • GitHub Copilot — AI pair programmer
  • Cline — VS Code AI agent
  • Windsurf — AI coding assistant
  • And any other agent supporting skills.sh

​
Support

Need help with the skill?

GitHub Issues

Report bugs and request features

Discord Community

Get help from our community