AI Agents Research Twenty Blockchains in Parallel
A shared evidence search can fan out across many targets before separate agents pursue the most relevant lines in depth.
Developed from a conversation between Pete Winn, Anthony and Andy David
Anthony used an agent workflow to investigate niche Bitcoin data across roughly 20 other blockchains. He was looking for evidence of stale Bitcoin blocks connected with those chains, a task that required scanning a wide field without letting any single chain consume the research effort too early.
The workflow gave every branch the same initial direction, then let agents cover the blockchains in parallel. That breadth-first pass established a consistent search across all the targets. From there, agents could go deeper on individual chains, turning the broad survey and the detailed investigations into stages of one research process.
This division of labour is where agent workflows can be useful. A person defines the goal and the evidence worth finding, while the workflow manages parallel branches, separate context windows and intermediate objectives. The agents can search widely and pursue each line tenaciously without requiring the person to specify every checkpoint or piece of scaffolding in advance.