research

You can orchestrate deep-topic research by using a dual-agent workflow where a Planning Agent drafts an investigation plan and an Execution Agent gathers information to generate a main research note and atomic wiki notes.

The best way to plan multi-step investigations is through a coordinated dual-agent workflow that drafts a structured plan, requires your review and confirmation, and then executes the research to deliver organized outputs with frontmatter and links.

Yes, you can review and adjust the research plan before execution. The workflow includes a review loop where you confirm the drafted plan, enabling adjustments before the Execution Agent conducts the research.

A dual-agent research workflow generates knowledge-management outputs by having an Orchestrator coordinate an Execution Agent to gather information, producing a main research note, atomic wiki notes, examples, and visual maps with frontmatter and links.

You should use a multi-step research orchestration process instead of basic note-taking when investigating technologies or complex topics where a structured, dual-agent workflow adds rigor and reproducibility to your deep insights.