Automate Research and Reporting

This bundle chains private web search into deep page scraping into structured knowledge storage into report generation. A research question goes to Brave Search for private, un-personalized results, then Firecrawl deep-crawls the promising sources - extracting content, following pagination, and exporting pages to Markdown. Notion holds the resulting knowledge base: structured notes with links between related findings. A Memory component persists context and accumulated knowledge across sessions, so a multi-day research effort does not start from zero each time.

On top of the retrieval stack sit purpose-built agents: a Research Agent that runs a multi-step search-and-synthesis pass on a topic and produces a structured report, a Market Research Agent for market size, trends, and key players, and a Competitor Analysis Agent that breaks down a competitor's products, pricing, positioning, and strengths/weaknesses. Two prompt templates - a blog post outline built from research findings, and a competitor battlecard for sales - turn the raw research into ready-to-use output.

The bundle maps its own pipeline as question to search to collection to analysis to synthesis to report, with each stage assigned to a specific component: Brave Search for the search step, Firecrawl for collection, Claude's own analysis for the analysis step, Memory for storage, and Notion for export. A worked example in the source asks for a 2024 no-code-platform market research pass: the top 10 players by revenue and users, key trends and growth drivers, a pricing-model comparison, a 3-year outlook, and named sources such as Gartner, Forrester, and CB Insights. The intended report format is fixed: an executive summary, methodology (sources and time period), numbered key findings with supporting evidence, an analysis section interpreting the data, recommendations, and an appendix of links/tables/charts.

Use it for structured research tasks - market sizing, competitor analysis, topic deep-dives - where you need both current web information and a persistent, organized knowledge base to build on over multiple sessions. It is not the right fit for a single quick lookup that does not need scraping, storage, or a formal report.

1. Formulate the research question.
2. Run the Research Agent for an initial information pass.
3. Go deeper on specific sources with Firecrawl.
4. Organize findings in Notion.
5. Produce a report with conclusions and recommendations.

Analysts, researchers, students, and anyone working with large volumes of information who needs it systematized into a reusable knowledge base rather than scattered notes.