---
title: "firecrawl"
description: "Search, scrape, crawl, and extract from the web with Firecrawl's official MCP server, with focused workflow skills."
canonical: https://agentpluginsdirectory.com/plugins/firecrawl
last-updated: 2026-09-22
---

# firecrawl
Search, scrape, crawl, and extract from the web with Firecrawl's official MCP server, with focused workflow skills.
- Slug: firecrawl
- Publisher: HiAI
- Repository: https://github.com/HiAi-gg/agent-plugins
- Manifest: plugins/firecrawl/plugin.json
- Version: 0.0.2
- License: MIT
- Category (editorial): research
- Skills: 4 (crawl-site, extract-structured-data, research-web, scrape-page)
- MCP servers: 1 (firecrawl)
- Stars: 1
- Repository created: 2026-08-08
- Repository last pushed: 2026-09-03
- Publisher type: Organization
- Listing: https://agentpluginsdirectory.com/plugins/firecrawl
- Schema: https://agent-plugins.org/schemas/1.0.0/plugin.schema.json

## What firecrawl does, in the publisher's words

Search, scrape, crawl, and extract from the web with Firecrawl's official MCP server, wrapped in focused workflow skills.

An Agent Plugin (spec 1.0.0).

A web research/extraction plugin: the official Firecrawl MCP server (firecrawl-mcp) plus skills for searching the web, scraping pages, crawling sites, and extracting structured data.

- agent-browser = interact with individual web applications/pages.
- Firecrawl = search/extract/crawl/research the web at scale.
- First-party upstream: Firecrawl's own MCP server, not a wrapper.
- Keyless tier: search, scrape, and interact work without an API key (rate-limited); no key is bundled.
- Read/extract focus: search, scrape, crawl, and extract are the workflows; no browser-control duplication with agent-browser.
- Server: firecrawl: stdio, npx -y firecrawl-mcp@3.23.4 (pinned; first-party firecrawl/firecrawl-mcp-server).

From the project README, punctuation lightly normalized. Full text: https://raw.githubusercontent.com/HiAi-gg/agent-plugins/HEAD/plugins/firecrawl/README.md

## Skills

- crawl-site: Crawl a website's pages with Firecrawl to inventory content, find broken links, or gather a corpus. Use when the agent needs to understand a site's structure, find all pages under a path, or collect a set of pages for analysis.
- extract-structured-data: Extract structured data from web pages or a set of pages with Firecrawl. Use when the agent needs fields, tables, or structured records from web content (pricing, listings, specs) rather than free text.
- research-web: Research a topic on the web with Firecrawl search: run searches, synthesize findings from results, and cite sources. Use when the agent needs current web information, competitive research, or to answer a question that requires up-to-date web sources.
- scrape-page: Scrape a web page into clean, agent-readable content with Firecrawl. Use when the agent needs the full content of a specific page (docs, articles, reference pages) for reading, extraction, or summarization.

Descriptions come from the frontmatter of each SKILL.md, punctuation lightly normalized.

## MCP servers

- firecrawl: transport: stdio; command: npx -y firecrawl-mcp@3.23.4

Read from the plugin's own mcp.json. Environment variable names only, never values.