Firecrawl Deep Research
Produce a sourced briefing on a market, product, or technical question by searching and scraping the web with Firecrawl instead of pasting raw SERP snippets into chat.
Overview
Firecrawl Deep Research is an agent skill most often used in Idea (also Validate, Build) that runs multi-angle Firecrawl search and scrape to deliver a sourced briefing on technical or market questions.
Install
npx skills add https://github.com/firecrawl/firecrawl-workflows --skill firecrawl-deep-researchWhat is this skill?
- Three depth tiers: quick (3–5 queries, 5–10 sources), thorough (5–10 queries, 15–25 sources), exhaustive (10+ queries, 2
- Firecrawl search-and-scrape via CLI with deduplication when search already returns full page content
- Onboarding caps at 1–3 questions only when topic, depth, or angle constraints are blocked
- Output is a sourced research report or briefing, not a dump of URLs
- Optional parallel sub-agents for collecting from multiple angles at once
- Quick tier: 3–5 search queries and 5–10 scraped sources
- Thorough tier: 5–10 queries and 15–25 sources
- Exhaustive tier: 10+ queries and 25+ sources including primary and contrarian sources
Adoption & trust: 11.8k installs on skills.sh; 29 GitHub stars; 2/3 security scanners passed (skills.sh audits).
What problem does it solve?
You need trustworthy, multi-source web evidence for a decision, but ad-hoc search in chat gives thin citations and inconsistent depth.
Who is it for?
Solo builders researching competitors, APIs, pricing narratives, or technical options who already use Firecrawl and want repeatable collection plans.
Skip if: Tasks that only need one URL summarized, offline-only corpora, or environments where you cannot provide a Firecrawl API key or outbound network access.
When should I use this skill?
User asks to research a topic, compare perspectives, produce a sourced briefing, investigate a technical or market question, or synthesize web evidence across many sources.
What do I get? / Deliverables
You get a structured research report backed by scraped sources at the depth tier you chose, ready to fold into validation docs or implementation planning.
- Sourced research report or briefing with cited web evidence
- Collection plan executed at quick, thorough, or exhaustive depth
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Canonical shelf is Idea because the skill’s default outcome is evidence-backed discovery before you commit to what to build or how to position it. Research subphase fits multi-query collection (quick 5–10 sources through exhaustive 25+), parallel sub-agents, and synthesis into a citable report—not one-off page fetches.
Where it fits
Run an exhaustive pass on primary sources and contrarian takes before you pick a niche.
Scrape competitor pricing and positioning pages into one sourced comparison memo.
Gather external evidence to narrow MVP scope and de-risk assumptions.
Investigate how vendors document APIs and rate limits across many official pages.
How it compares
Use instead of asking the agent to “search the web” without a scrape plan or source budget.
Common Questions / FAQ
Who is firecrawl-deep-research for?
Indie builders and agent users who need sourced briefings—market, product, or technical—and want Firecrawl-driven search and scrape rather than hallucinated summaries.
When should I use firecrawl-deep-research?
In Idea research when scoping an opportunity; in Validate when comparing positioning or pricing evidence; in Build when investigating integrations or vendor docs across many pages.
Is firecrawl-deep-research safe to install?
It requires a Firecrawl API key and network calls to third-party sites; review the Security Audits panel on this Prism page and rotate keys if you share agent logs.
SKILL.md
READMESKILL.md - Firecrawl Deep Research
# Firecrawl Deep Research Use this when the user wants a sourced research report, not raw search results. ## Onboarding Interview Infer the topic, depth, and output format from context. If the topic is clear, proceed immediately. Ask at most 1-3 concise questions only if blocked, such as the research topic, required depth, or a critical angle/source constraint. ## Firecrawl Collection Plan Use Firecrawl search and scrape through the CLI or equivalent tool surface. - Quick: search 3-5 queries and scrape 5-10 high-quality sources. - Thorough: search 5-10 queries from different angles and scrape 15-25 sources. - Exhaustive: search 10+ queries and scrape 25+ sources, including primary sources, research papers, expert views, and contrarian sources. Avoid re-scraping URLs already returned with full content from a search-with-scrape result. ## Parallel Work If appropriate, use sub-agents or equivalent parallel task runners by research angle: - overview and definitions - technical or implementation details - market and industry context - contrarian views, risks, and limitations - primary sources and official docs Each researcher should return claims, source URLs, source quality notes, and uncertainty. ## Final Deliverable Default structure: ```markdown # Deep Research: [Topic] ## Executive Summary [2-3 paragraphs] ## Key Findings [Numbered findings with source links] ## Detailed Analysis [Themes, evidence, and synthesis] ## Contrarian Views And Risks [Counterarguments, limitations, failure modes] ## Open Questions [What remains uncertain] ## Sources [Every URL used with a one-line note] ## Rerun Inputs workflow: firecrawl-deep-research topic: [topic] depth: [quick/thorough/exhaustive] output: [markdown/json/brief] ``` ## Quality Bar - Cite sources for factual claims. - Prefer primary sources when available. - Flag uncertainty and conflicting evidence. - Synthesize instead of listing scrape summaries.