One model researching a broad question gets shallow fast; it cannot read widely and fact-check itself at the same time. This crew splits the job across specialists.
sonnet1 weekCrewAITavilyFirecrawlOpenAI
Claude
81ROI
79Scale
$4k90Saved
ROI for
README.md
Why this subagent
One model researching a broad question gets shallow fast; it cannot read widely and fact-check itself at the same time. This crew splits the job across specialists.
A lead agent breaks the question into sub-topics, scrapers gather sources in parallel, and synthesis and verifier agents merge and check the findings. You get a cited research brief where every claim traces back to a source, not a confident guess.
How it runs
Used at step 01 to kick off the pipeline.
Write
Used at step 01 to kick off the pipeline.
WebFetch
Used at step 01 to kick off the pipeline.
WebSearch
Used at step 01 to kick off the pipeline.
Scraper agents gather sources for their sub-topics in parallel while a synthesis agent waits for their output.
pending
The synthesis agent merges findings into one draft and a verifier agent checks each claim against its source.
pending
The lead agent resolves conflicts, drops unsupported claims, and returns a cited brief to Notion.
pending
Sample output
json
// Sample output
// (generated when the pipeline finishes)
As the lead agent, decompose the research question into sub-topics, dispatch parallel scrapers, then have synthesis and verification agents produce a single brief where every claim carries a source link.
Unlock the rest
The full agent definition, install snippet, and starter task are gated for community members.
Members get the full `.md` agent file, the npm / pnpm install one-liners, a starter prompt that we've tuned against real runs, and the open-source repo when this automation ships there. One email, magic link, done.