A workspace for preparing tabletop adventures and running sessions. Review imported suggestions against their source, then bring confirmed content into Run.
Read the case study
This demo uses original fictional content and deterministic mock extraction. Assessment scores are preset, not measured AI accuracy.
