globalpercent

Based on the available evidence, globalpercent is best judged as an open-source data-access and panel-style AI skill, not a broadly validated production research product. The GitHub repository supports that it provides a “global macro probability panel” and a zero-auth pipeline over Polymarket and Kalshi, but the evidence is mostly limited to the repo page itself, so adoption and reliability should be assessed conservatively.

In practice, it looks like a macro-probability data layer for AI coding assistants: something a model, agent, or script can call to produce structured panels, summaries, or macro context from prediction-market signals. It is not a general-purpose foundation model, and it is not a full quantitative trading platform. A better analogy is a lightweight market-data skill or wrapper for AI agents and coding workflows.

On barrier and cost, the evidence only directly supports the repository claims such as “open source” and “zero-auth data pipeline.” That suggests lower setup friction than tools requiring API keys, but it does not prove stable, long-term, or cost-free service.