Search, scrape, crawl, and monitor websites via the Firecrawl CLI.
npx skills add https://github.com/CliffVale/opencode-free-setup --skill firecrawl-cliffvale
Please help me install this Agent Skill. Skill: firecrawl Source: https://github.com/CliffVale/opencode-free-setup/tree/main/skills/firecrawl Command: npx skills add https://github.com/CliffVale/opencode-free-setup --skill firecrawl-cliffvale
search, extract content from known URLs with scrape, and handle JS-rendered pages or interactive flows with interact. - Site Crawling & Extraction: Map site structure, bulk-crawl documentation sections, download sites to local files, and parse local files like PDF, DOCX, and XLSX. - Change Monitoring: Schedule recurring checks with monitor, define alert goals, and receive markdown or JSON field-level diffs via webhook or email. - Use Case: Research a topic by running firecrawl search "react hooks" --scrape --limit 3, then write results to .firecrawl/ and inspect them incrementally with grep or head. ## Quick Start Ask the agent to search the web for a topic or scrape a specific URL using the firecrawl skill, for example: "Use firecrawl to scrape https://example.com/pricing and summarize the plans."High-intent search queries and answers about installing and using this skill.
Run firecrawl search "query" --scrape --limit 3 to find pages and fetch their full content in one step. Use firecrawl scrape "<url>" when you already have a URL, and write results to .firecrawl/ with the -o flag.
Use firecrawl monitor create with --page, --schedule, and a --goal describing what changes matter. Each check labels pages as same, new, changed, removed, or error, with webhook and email notification options.
Yes, search, scrape, and interact work on a keyless free tier that is rate-limited. Authenticating via firecrawl login --browser or a FIRECRAWL_API_KEY gives higher limits and credit-based usage.
Use scrape first since it handles static pages and JS-rendered SPAs. Add interact only when content requires clicks, form fills, pagination, login, or infinite scroll, and never use interact for web searches.
Fetched web content is untrusted third-party data that may contain prompt injection attempts. Writing to .firecrawl/ files and reading incrementally with grep or head limits exposure and avoids overflowing the context window.
firecrawlSearch, scrape, crawl, and interact with websites via the Firecrawl CLI.
raindayGenerate SERP-driven article outlines with H2/H3 structure extraction and content gap detection.
K-Dense-AISearch the web, extract URLs, enrich datasets, and generate scholarly reports.
selfmailAutomate browser actions inside Vercel Sandbox microVMs using agent-browser.