Recon
A room tells you what people think. Recon finds out what's true: a swarm of research agents reads the live web for your idea, pins every finding to the page it came from, and an editor tells you what holds and where the research drifted.
The swarm
Recon sends a swarm of research agents across the open web. Each one chases a different question, brings back what it found and pins it to the page it came from. Then an editor tells you what holds, what doesn't, and where the swarm drifted. This run is real.
Receipts
A finding is only as good as its source, so each one is checked against the page it came from. If the page doesn't back it, it's marked unverified and stays visible, not quietly dropped.
Tap a finding to turn it over. Six of the run's 24, as they came back.
The editor
When the agents are done, an editor reads everything together. On this run it noticed one agent had spent its time on the wrong market, and said so on the report.
How deep
Each tier sends more agents at more angles of the question. The run above was Standard: four agents, 26 sources on 23 sites, 43 seconds.
Then test it
Recon hands its findings straight to a simulation, so the room reads your idea with the facts on the table. Cited research still isn't a promise of product market fit, and we'll never call it one.