Hi Vladimir, The trends such as PR count, commit activity, reviewer diversity, mentor sign-off rate, release cadence, and mailing-list participation - are already being computed from real data across 3m, 6m, and 12m windows. The model is working from actual measured signals, not guessing.
What the prompt does not yet do is define explicit thresholds for when a signal becomes a concern. That could be improved, but I’m not sure there are universal rules that could apply to all communities. This is an initial experiment and is very much open to improvements. Kind regards, Justin --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
