CausalCanvas

CausalCanvas is a web app (with optional desktop wrapper) for building and stress-testing causal loop diagrams and stock-and-flow models without needing a full-blown simulation suite. Users sketch variables, links, delays, and feedback loops, then run lightweight scenario tests with simple parameter ranges to see which loops dominate and where interventions backfire. The app includes a structured “assumption ledger” so every link has a source, confidence score, and owner—useful for teams arguing over what’s true. An AI copilot can translate meeting notes into an initial map, suggest missing variables, and flag common modeling mistakes (circular definitions, inconsistent units, hidden exogenous drivers). Export to PDF/PNG and to formats compatible with common modeling tools. This is not a magic prediction engine; it’s a practical thinking tool that makes system structure explicit and reviewable.