firecrawl-build-search

Name: firecrawl-build-search
Availability: InStock
Author: Firecrawl

Discover web pages and sources matching a search query for downstream extraction or analysis.

200k|10.5k|Updated Apr 15, 2024

Search Tags:#search #firecrawl #query #web discovery #source ranking

Ready to run

npx skills add https://github.com/firecrawl/firecrawl --skill firecrawl-build-search-firecrawl

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: firecrawl-build-search
Source: https://github.com/firecrawl/firecrawl/tree/main/skills/firecrawl-build-search
Command: npx skills add https://github.com/firecrawl/firecrawl --skill firecrawl-build-search-firecrawl

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When your application starts with a user question or research topic rather than a known URL, you need to discover relevant web sources first. This Skill integrates Firecrawl's `/search` endpoint to find and rank web pages before extracting their content. ## Core Features & Use Cases - Query-to-URL Discovery: Convert search queries into a ranked list of relevant web pages with snippets. - Source Selection: Identify authoritative sources for answer generation, competitive analysis, or research workflows. - Optional Hydration: Retrieve full page content for selected results without separate scraping calls. - Use Case: A financial research assistant receives "What are the latest funding rounds in biotech?" and uses this Skill to search the web, identify investor news sites and company announcements, then optionally extract full articles for summarization. ## Quick Start Use the Firecrawl search endpoint with your query to discover web pages, then pass the most relevant URLs to the scrape or interact skills for deeper extraction.

Frequently Asked Questions about firecrawl-build-search

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I search the web and get URLs before scraping with Firecrawl?▼

Use the /search endpoint with your query string to retrieve ranked results containing URLs, titles, and snippets. Pass the selected URLs to /scrape or /interact for full content extraction without extra API calls.

When should I use Firecrawl search versus building a scraper for a specific site?▼

Use /search when discovering sources is part of the workflow and URLs are unknown. If you already know the target URL or need content from a fixed set of sites, use /scrape directly to save latency and cost.

Can Firecrawl search retrieve full page content in a single request?▼

Yes, optional hydration can fetch content for selected results, but search results primarily return snippets and metadata. For consistent full-text extraction, combine search discovery with separate /scrape calls.

How do I search for published research papers with Firecrawl?▼

The /search endpoint with categories=['research'] filters to research-affiliated sites but does not query abstract databases. For biomedical literature (PubMed, bioRxiv, medRxiv) or arXiv preprints, use the dedicated research index skill instead.

Does Firecrawl search support filtering by category or date?▼

The /search endpoint supports category filtering (web, research, developer) to focus results on specific domains. Refer to source-of-truth documentation for your language for parameter details and current filtering options.

What formats does Firecrawl search return?▼

Search results return structured JSON with URLs, titles, snippets, and metadata. Results are optimized for LLM processing and agent workflows to minimize token consumption.