How this skill is triggered — by the user, by Claude, or both
Slash command
/crawl4ai:setupThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Set up the Crawl4AI plugin for Claude Code.
Set up the Crawl4AI plugin for Claude Code.
Guide the user through setup:
Check current config:
Read ~/.crawl4ai/claude_config.json if it exists. Report current mode and whether API key is set.
Choose mode: Ask the user:
pip install crawl4ai && crawl4ai-setup.Configure based on choice:
For cloud mode:
sk_live_ or sk_test_)~/.crawl4ai/claude_config.json:
{"mode": "cloud", "api_key": "sk_live_xxx"}
CRAWL4AI_API_KEY env varFor local mode:
python3 -c "import crawl4ai; print(crawl4ai.__version__)"pip install crawl4ai && crawl4ai-setup{"mode": "local"}Verify setup:
Use the crawl MCP tool to crawl https://example.com. Confirm it returns markdown content.
Report status: Tell the user their setup is complete and which mode is active.
npx claudepluginhub unclecode/crawl4ai-cloud-sdk --plugin crawl4aiScrapes URLs to markdown/HTML/JSON, crawls websites for multi-page extraction, searches the web, maps sites, and extracts structured data using Firecrawl MCP tools.
Crawls websites and extracts content from multiple pages via the Tavily CLI. Supports depth/breadth control, path filtering, semantic instructions, and saving pages as local markdown files.
Reference for building production-ready crw integrations: verb selection, CLI/MCP/REST call surfaces, post-filtering, context-window hygiene, Hybrid RAG patterns, and operational pitfalls.