crawl4ai-claude-plugin

Crawl4AI Claude Code Plugin

Web crawling and LinkedIn prospecting plugins for Claude Code, powered by Crawl4AI.

Plugins

Plugin	Description	Commands
crawl4ai	Crawl websites → structured markdown	/crawl
linkedin-prospect	LinkedIn prospecting pipeline	/li-discover, /li-insights, /li-prospect

Installation

From GitHub (any Claude setup)

claude plugin add github:TheophilusChinomona/crawl4ai-claude-plugin

From local directory (development)

claude plugin add directory:/path/to/crawl4ai-claude-plugin

Prerequisites

• Python 3.10+
• pip install crawl4ai
• crawl4ai-setup (installs headless browsers)

Additional for linkedin-prospect

• pip install litellm sentence-transformers pandas numpy scikit-learn rich
• A LinkedIn browser profile created via crwl profiles

Quick Start

# Single page crawl
/crawl https://example.com

# Deep crawl with pruned markdown
/crawl https://docs.python.org --depth 2 --max-pages 50 --fit

# Batch fetch from a URL list file
/crawl --url-list urls.txt --fit

# Full site crawl
/crawl https://docs.example.com --all

# LinkedIn prospecting
/li-prospect "health insurance management" --geo 102713980 --max-companies 10

Output

crawl4ai

.crawled/
  _index.md                    # Master index
  example.com/
    page-title.md              # Markdown with YAML frontmatter
    docs/
      getting-started.md

linkedin-prospect

linkedin-data/
  companies.jsonl              # Scraped companies
  people.jsonl                 # Scraped people
  insights/
    decision_makers.csv        # Top prospects
    graph_view.html            # Interactive visualization

License

MIT