external-context

You can gather sourced web research by decomposing a coding architecture question into multiple independent search facets and spawning parallel agents to collect official documentation, examples, and references. This approach compiles a cited brief with key findings and consolidated URLs from multiple concurrent perspectives.

The best way to compare libraries is through query decomposition into 2–5 independent search facets, enabling parallel document-specialist lookups across different angles. This concurrent perspective gathering produces a structured summary with detailed per-facet results and a final list of source URLs.

You get cited best-practice lookups by orchestrating parallel web lookups that spawn up to five concurrent agents to gather official docs and references per facet. The agents synthesize results into a structured summary with key findings, per-facet details, and a consolidated sources list.

Yes, you can use parallel agents to research multiple independent search facets simultaneously, with a maximum of five concurrent agents. Each agent acts as a document specialist gathering official documentation, examples, and references for its assigned facet, which are then compiled into a single cited synthesis output.

Yes, parallel external research works effectively for documentation-driven investigations by decomposing queries into 2–5 independent search facets and collecting documentation from the web. It applies to best-practice lookups, library comparisons, and architecture questions that benefit from multiple concurrent perspectives.

Facet decomposition for external web research is limited to a maximum of five concurrent parallel agents and decomposes queries into 2–5 independent search facets. It is specifically designed for coding, architecture, library comparisons, best-practice lookups, and documentation-driven investigations rather than general browsing.