in_pool)
·
Off-pool = not on manifest loot-pool export (in_pool); quests/vendors/rewards still possible — see validation FYI + Sources
·
Inv tags = composition basetags + per-slot tag chain from exported inv data; failures show as Inv tags: … in validation
·
NCS = raw catalog slot (no pool filter)
·
Fail (data) = breaks a rule in the validation panel (not “illegal in-game”)
Paste an @U serial. The STX decoder (same as BL4 Bulk Deserializer) resolves resolvedParts, maps names to manifest slots (best-effort), then runs the same data checks as manual selection — including inv composition tags when inv_comp_tag_data.js is loaded. Empty or partial mapping still runs checks on whatever matched.
Uses manifest / NCS / schedules / weights / stats and inv tag composition rules; panel lines starting with Inv tags: come from the tag graph.
parts_stats_data.js) · WEAPON_STATS_DATA · LootSchedule & weapon_part_weight_table · inv tag rules (inv_comp_tag_data.js + tag-comp-validation.js)
LootSchedule / weapon_part_weight_table come from extracts; mapping a chosen part to a schedule row still uses heuristics we maintain.
Status labels (OK / Fail / Uncertain) describe agreement with our tables and rules, not a private “legit” verdict from the game. Stats by idRaw / Stats known come only from PARTS_STATS_DATA here. For full part lookup (same STX idRaw layer), use Advanced Part Search in the main toolbox. For many serials at once (same checks + export alignment flags), use Bulk serial validator.
Slot names are checked against NCS (with manifest aliases: mag/magazine, shield/hyperion_secondary_acc, multi/tediore_acc). Element overlap across slots is a consistency heuristic we chose (same element token in two slots); the engine may differ. Rainbow / multi-element single parts are skipped for overlap checks.