name: firecrawl-deep-research\ndescription: |\nProduce an intensive, cited analytical report: executive summary, multi-angle\nfindings, contrarian views, open questions, and full sources. Use only when the\nuser needs rigorous synthesis of a complex topic (scientific, technical, policy,\nor market-analytical) that cannot be answered with a short search, and wants\na formal written report, not a recommendation list.
\nDo not use for product picks, top-N lists, quick lookups, or routine \"find out\nabout X\" tasks. If the request does not clearly need this kind of report, do\nnot use this skill.
\nDo not use for a literature review over published papers. This skill collects\nevidence from the open web. A request for the literature on a biomedical,\nclinical, life-science, or other scientific topic — papers, studies, trials,\npreprints — belongs to firecrawl-research-papers, which queries Firecrawl's\npaper index (PubMed, bioRxiv, medRxiv, arXiv) instead of searching websites.\nlicense: ISC\nmetadata:\nauthor: firecrawl\nversion: \"0.1.0\"\nhomepage: https://www.firecrawl.dev\nsource: https://github.com/firecrawl/firecrawl-workflows\ninputs:
\nUse this only for report-scale research: a rigorous, cited synthesis the user\nexplicitly wants delivered as a formal written report. If the request is a\nproduct pick, a top-N list, a quick lookup, or anything answerable with a short\nsearch, stop; do not use this skill, let the request be handled the standard\nway.
\nThis skill gathers its evidence from the open web. If the evidence base is the\npublished literature — a literature review, or a biomedical, clinical, or other\nscientific topic where the answer lives in papers — use\nfirecrawl-research-papers instead; it\nqueries Firecrawl's paper index rather than searching websites.
\nInfer the topic and output format from context. Before starting, unless already specified, always ask one short question to define the scope:
\n\n\n\"How long do you want this research task to run?\"
\n
Map the answer to a depth tier in the Collection Plan below:
\nIf the topic itself is unclear, you may ask at most 1-2 additional concise\nquestions (topic, or a critical angle/source constraint). Otherwise proceed once\nthe runtime is set.
\nUse Firecrawl search and scrape through the CLI or equivalent tool surface. Match\ndepth to the runtime the user chose during onboarding.
\nAvoid re-scraping URLs already returned with full content from a search-with-scrape result.
\nSearch and scrape reach web pages. They do not query Firecrawl's research paper\nindex, which holds paper abstracts with full text reachable per paper — largely\nbiomedical and life-science literature from PubMed, bioRxiv, and medRxiv, plus\narXiv preprints in CS, physics, and math.
\nHand off to firecrawl-research-papers\nwhen the report's evidence base is the published literature — a biomedical,\nclinical, drug, gene, disease, epidemiology, or public-health topic, or any\nrequest phrased as a literature review, systematic review, or survey of studies.\nThat skill uses firecrawl_research_* (MCP) / firecrawl research (CLI) to\nsearch abstracts, expand to related papers, and verify claims inside a paper\nbody — none of which plain search and scrape can do.
If the report needs both — the literature and market, policy, or news context —\nrun the paper work through that skill and keep the web collection above for the\nrest, then synthesize here.
\nNote that passing categories: [\"research\"] to Firecrawl search does not query\nthe paper index either. It filters an ordinary web search to research-affiliated\nwebsites — the list includes PubMed, bioRxiv, medRxiv, arXiv, and publisher\nsites — and returns their web pages, not the paper records behind them.
If appropriate, use sub-agents or equivalent parallel task runners by research angle:
\nEach researcher should return claims, source URLs, source quality notes, and uncertainty.
\nDefault structure:
\n# Deep Research: [Topic]\n\n## Executive Summary\n[2-3 paragraphs]\n\n## Key Findings\n[Numbered findings with source links]\n\n## Detailed Analysis\n[Themes, evidence, and synthesis]\n\n## Contrarian Views And Risks\n[Counterarguments, limitations, failure modes]\n\n## Open Questions\n[What remains uncertain]\n\n## Sources\n[Every URL used with a one-line note]\n\n## Rerun Inputs\nworkflow: firecrawl-deep-research\ntopic: [topic]\ndepth: [quick/thorough/exhaustive]\noutput: [markdown/json/brief]\n