deep-research

Complex research objectives are decomposed by breaking them down into structured sub-questions and executable tasks, which are then delegated to specialized subagents for parallel domain-specific analysis across large contexts.

Large-context document analysis is performed using long-context reasoning to process extensive documentation and search results, aggregating findings and synthesizing insights into coherent, well-supported analyses.

Task orchestration delegates tasks to specialized subagents for parallel and domain-specific analysis, coordinating their execution to aggregate findings and synthesize insights for iterative research workflows.

Yes, executing multi-step research planning and analysis requires a Crafted MCP server and an API key to orchestrate specialized subagents and process large-context reasoning.

Yes, it processes extensive documentation using long-context reasoning, aggregating findings and synthesizing insights into coherent reports and well-supported recommendations for iterative research workflows.

Avoid orchestrated subagents for simple, single-step queries that do not require planning, decomposition, or long-context reasoning across various tools and files, as the orchestration overhead outweighs the benefits.