crystal-ball

Based on the available evidence, crystal-ball is best judged as a promising but thinly evidenced open-source intelligence workbench. The current support is enough to confirm the stated product direction—Tauri 2 desktop app, 3D globe, live panels, MCP server, and an explainable intelligence layer—but not enough to claim broad validation. The visible GitHub numbers, 6 stars and 2 forks, are only proof of early attention, not proof that it works well in real production use.

In practical terms, it looks more like an OSINT/analysis desktop workbench prototype than a general BI tool, a plain mapping SDK, or a mature enterprise risk platform. A better comparison is a local desktop app that bundles Cesium-style geospatial visualization, intelligence panels, and Claude Code MCP connectivity, then layers on explainable outputs such as truth scoring, evidence graphing, compound risk, and forecast calibration.

On cost and adoption friction, the evidence only supports that the code is open and publicly available. There is no official pricing, hosted plan, API fee sheet, or support policy in the provided sources, so total cost of ownership should be treated conservatively as unknown.