sn-deep-research

To generate an evidence-backed deep research report, you orchestrate multi-dimension data collection, cross-source verification, and synthesis. The workflow records your request, plans evidence gathering by dimension, tracks uncertainties, and drafts a final report with strict progression guards.

The best way to structure market research into verifiable sub-reports is by breaking down the topic into dimensions. Each dimension gets key questions, evidence notes, uncertainty tracking, and completion criteria before synthesis and final report drafting occur.

Yes, you can resume an interrupted deep research workflow by pointing to the same report directory. The process supports staged artifact persistence, saving files like request.md, plan.json, and sub_reports to seamlessly continue from the last completed step.

Yes, generating an investment-style memo requires a pre-flight web search probe. This availability check ensures the environment can perform cross-source verification and fact-checking before orchestrating the multi-angle data collection and synthesis phases.

The research synthesis process prevents skipping to final conclusions to ensure evidence sufficiency. Strict progression guards mandate that fact-checking and multi-source validation are fully completed across all dimensions before the final report is written.

You need cross-source verification for policy and pricing research when dealing with multi-angle comparisons or timely fact-checking. This ensures your evidence-backed report reliably synthesizes regulations, market data, and product details instead of relying on a single viewpoint.