arXiv cs.CLPaper
Terminal-Universe: Turning Agent Trajectories into Scalable Terminal Environments
Agent training has been bottlenecked by environment scarcity; this flips that by recycling trajectories into new environments. If you're doing post-training on code agents, this significantly lowers the cost of creating diverse interaction data. The method depends on trajectory completeness, but the idea scales to any agent domain that logs state changes.