# Gaps — conflict & fragility SEA

## Intentional empties
- **ACLED events / fatalities** — all 11 empty. No verified SEA-11 country totals locked without paywall/login. Do not invent.
- **GPI Brunei** — Brunei is absent from GPI 2025 country coverage. Empty cells only.

## conflict_flag derivation (transparent)
| Value | Rule |
|-------|------|
| `major_active` | Myanmar — civil war / major armed conflict post-2021 coup; FSI Alert ≥100; GPI Asia-Pacific floor |
| `subnational` | GPI 2025 conflict dyads name a government–nonstate pair inside the country (PH, TH, ID) |
| `low_or_none` | No major/subnational dyad in GPI 2025 conflict tables for that country |

This is a **label**, not a continuous intensity score. Never average it with FSI or GPI.

## Caveats
- FSI rank **1 = most fragile**; GPI rank **1 = most peaceful** — opposite orientation. Keep panels separate.
- WGI PV is a perception estimate, not event counts.
- Do not duplicate INFORM overall risk from `governance_risk`.
- Myanmar macro series elsewhere carry conflict-era uncertainty badges — keep them there, not here.

## Completeness
All 11 have WGI PV 2024 + FSI 2024 score/rank/band. GPI 2025 for 10/11 (BRN GAP). conflict_flag for all 11.