PWAS looks for software worth building and shows its reasoning. Ideas are cheap; evidence is what counts.
PWAS reads public sources through their official APIs and keeps the posts that point to a real need.
Every fact is stored with its source and a label that says how it is known: verified, public claim, estimate, AI inference, community, or unknown.
The PWAS score says how attractive an opportunity looks. Confidence says how much of that is backed by evidence. They are always shown together.
A model sorts each factor into a rubric bucket and cites the evidence it used. Code turns buckets into numbers, so any score can be recomputed and explained.
Members vote BUILD or PASS and challenge the evidence. Votes alone never trigger a build: the score, the evidence and a legal check must also hold.
The long road from an opportunity to an exit. The first stages are live; the rest open phase by phase.