# FLR001-LAB1 — Dependency Divergence Diagnosis Record v1

Use the eight deterministic synthetic fixtures in `fixtures/cases.json`.

For each case:
1. identify the evidence layer being compared;
2. separate declared state from observed/resolved/executed state;
3. confirm whether the symptom is reproducible;
4. locate the earliest divergence that can explain the failure;
5. record negative evidence that weakens at least one alternative;
6. choose reversible containment before broader repair;
7. require an independent rerun plus recurrence test for repair closure; and
8. preserve missing runtime/platform facts as OPEN rather than guessing.

Implement `build_record(case)` in `starter.py`. The learner files contain no hidden expected classifications. Everything is local, synthetic, and offline.