Agents Lookup

Use cases / Research

Broad research questions split across parallel subagents

A lead agent splits a broad question into parts, sends subagents to research each in parallel, and merges their findings.

DifficultyAdvanced
Setup time1-2 weeks
CategoryResearch

The use case

Questions like 'list every company in this market and what they offer' are too broad for one search thread. Anthropic described its multi-agent research system: a lead agent plans the work, spawns subagents that each search one part in parallel, then combines what they return and a separate step attaches citations. They found this helps on breadth-first questions but uses many more tokens. Teams building internal research tools can copy the pattern. A researcher still reviews the merged output for gaps and duplicates and checks citations on claims that matter.

Example workflow

  1. Lead agent breaks the question into parts
  2. Subagents research each part in parallel
  3. Lead agent merges and deduplicates results
  4. Citation step links claims to sources
  5. Researcher reviews gaps and checks key citations

Tools

Risks and guardrails

Sources

Written by Agents Lookup. Credits and official pages for the tools this use case names: