Problem
A solo builder sees plenty of trends and pain points every day, but the stretch from "this looks like a signal" to "this is worth building" — filtering, competitive research, diagnosis, validation — is a path almost nobody has actually built.
Decisions
Filtering, signal research, opportunity analysis, an AI-tone editorial pass, a competitor knowledge base, case output — each of the six steps stands alone and can be rerun on its own, without starting over.
Outcome
The local workspace contains versioned prompts, real and synthetic run records, competitor matrices, case packages, and a static review site.
Reflection
Finding an opportunity was never about a flash of inspiration — it's putting the evidence, what's still uncertain, and what to verify next on the same table at the same time.