deep-research

Multi-source web research uses a Graph-of-Thoughts pipeline to plan, execute, and validate investigations. It performs landscape scans, triangulates evidence across sources, and labels claims with confidence levels to generate a structured markdown report.

Triangulation in claim verification cross-references evidence across multiple independent sources to validate findings. It labels claims with C1, C2, or C3 confidence levels, ensuring that conclusions drawn from fast-changing topics are robust and reliable.

To run a Red Team critique during an investigation, trigger the research pipeline to execute its self-critique phase. This built-in step challenges the gathered evidence, performs gap analysis, and tests hypotheses against alternative viewpoints to identify vulnerabilities.

Yes, you can generate cross-industry insights by activating Creative Mode within the research pipeline. This mode applies hypothesis-testing and a Graph-of-Thoughts structure to draw parallels from unrelated fields, yielding innovative solutions for complex problems.

The best way to structure complex investigations is using a tiered pipeline with Quick, Standard, Deep, and Exhaustive modes. This approach systematically moves through scoping, planning, retrieving, and validating data to produce comprehensive research reports.

Yes, structured multi-source research operates without external dependencies by relying on standard browsing tools. It intrinsically manages the landscape scan, evidence retrieval, and gap analysis processes to output a complete markdown research report.