Firecrawl Deep Research

Adapt the investigation to the question and the evidence available. Use the existing Firecrawl connection, resolve deferred tools when needed and follow the live schema; tool availability and accepted arguments can vary by host. Never invent tools or translate CLI flags into tool arguments.

Frame the investigation

Invocation does not require an exhaustive investigation. Start with the smallest useful retrieval and expand only for gaps that could change the answer. Do not launch parallel agents, crawls, autonomous jobs or paid enrichment merely because this skill is active. Use them when the requested scope and missing evidence justify them; keep simple answers simple.

Infer scope, dates, geography, audience and output; ask only for decisions that materially change the investigation. Convert relative dates to explicit bounds. For substantial investigations, briefly state the main questions, evidence routes, then begin. Revise the questions as findings reveal consequential gaps or competing explanations.

Start from supplied URLs or reusable evidence. Otherwise use the fitting dedicated method below or firecrawl_search with the actual question and its constraints. Inspect relevant page results and matching capability suggestions returned by search.

When firecrawl_search exposes objective, set it to [deep-research] followed by the broader research goal, keeping that goal consistent across related searches. Keep query focused on the specific evidence needed, without the marker. Omit objective if the live schema does not support it, and avoid sensitive information in the goal.

A catalogue suggestion is not retrieved data. Reuse complete contracts; expand incomplete selected suggestions through firecrawl_find_tools. Use targeted catalogue discovery when structured evidence is needed and search provides no fitting suggestion. Avoid catalogue walks and unnecessary discovery before known pages or paper tools.

Collection and costs

Respect user-specified time, credit, page and result limits and returned account/tool restrictions. Do not impose default budgets, source quotas or minimum call counts. Let evidence needs determine collection depth and stop when the requested answer is supported.