ALLOWLISTED DURABLE SOURCE

003-public-record.md

docs/builds/003-public-record.md

This is the retained record itself—not a chat summary. It is exposed because it materially documents the research, method, evidence, audit, production, planning or release history of The 100.

# Build 003 — Public Record

## Making
003 began with a deliberately inspectable decision-gap heuristic rather than pretending an AI model could discover product-market fit. The product separates situation, unresolved decision, existing information substitutes and four structural signals. Optional browser-local AI can challenge hypotheses but is never treated as demand evidence.

## Method
1. Name the decision people still cannot make.
2. Inventory what information they already have.
3. Test whether the decision is recurring, consequential, ambiguous and fragmented.
4. Generate bounded hypotheses about decision infrastructure, not claims of demand.
5. Validate with people/evidence later; do not convert a heuristic score into market truth.

## Evidence + limitations
The four dimensions and current score formula are a product-discovery heuristic created for this build. They are not a validated market-demand scale and are not attributed to an external methodology. The score is useful only as an explicit, challengeable prioritization device. Public-source evidence/search is intentionally assigned to Build 004's intelligence/retrieval infrastructure rather than duplicated here.

## Privacy/support
Core scoring and hypotheses are deterministic and client-side. Browser-local AI is optional and feature-detected. No server-side market research or external evidence retrieval is claimed in 003. Input length/resource limits, cold-start/model failure and device support remain acceptance items before release certification.