DNF90 服务端与启动器完整源码（含 GM 工具功能增强）

服务端（Go）：
- GM 塞背包/带属性装备发放：在线角色推送 op14 增量物品行，背包实时刷新，无需回角色选择界面
- GM 改金币/改等级：去掉在线拒绝，落库后推送 op14 钱包/op37 等级状态，在线实时刷新
- 等级端点统一切换到新版 GMSetCharacterLevel（自带在线刷新）
- 无限/锁死疲劳模式 + 燃烧疲劳：每房间消耗 2 点、打怪与通关经验双倍
- 燃烧疲劳开关时给所有在线角色推送系统提示（op12 mode 0）
- 深渊史诗掉率：GM 全局倍率 + 城镇增益药剂窗口倍率，真实接入掉落结算

启动器（Python/JS）：
- 标题栏自绘、窗口 1204x717、禁止拖拽改大小、仅最小化/关闭
- GM 工具：疲劳与燃烧、深渊史诗掉率、拍卖行上架（单一定价对齐）
- 拍卖行上架前端对齐服务端单一价格语义，修复 400 报错

server/internal/modules/dnf/progression/skill_points.go

@@ -0,0 +1,138 @@
+package progression
+
+import (
+	"fmt"
+	"math"
+
+	dnfrepo "longheng.io/server/internal/modules/dnf/repository"
+)
+
+type SkillPointAdvance struct {
+	Previous dnfrepo.SkillPointState
+	New      dnfrepo.SkillPointState
+	SPGain   int
+	TPGain   int
+}
+
+func (t *Tables) SkillPointsAtLevel(level int) (int, bool) {
+	if t == nil {
+		return 0, false
+	}
+	value, ok := t.spByLevel[level]
+	return value, ok
+}
+
+func (t *Tables) TechniquePointsAtLevel(level int) (int, bool) {
+	if t == nil {
+		return 0, false
+	}
+	value, ok := t.tpByLevel[level]
+	return value, ok
+}
+
+// TotalSkillPoints returns the cumulative PVF SP through level. A missing row
+// is an error rather than an implicit zero because silently skipping a runtime
+// PVF level would permanently underpay a character.
+func (t *Tables) TotalSkillPoints(level int) (int, error) {
+	if t == nil || level <= 0 {
+		return 0, fmt.Errorf("%w: SP level=%d", ErrLevelOutOfRange, level)
+	}
+	total := 0
+	for current := 1; current <= level; current++ {
+		award, ok := t.spByLevel[current]
+		if !ok {
+			return 0, fmt.Errorf("%w: %s missing level=%d", ErrTableMalformed, SkillPointTablePath, current)
+		}
+		if award > math.MaxInt-total {
+			return 0, fmt.Errorf("%w: %s cumulative SP overflow at level=%d", ErrTableMalformed, SkillPointTablePath, current)
+		}
+		total += award
+	}
+	return total, nil
+}
+
+// TotalTechniquePoints returns the cumulative PVF TP through level. Levels
+// below the first TP row award zero; once the table starts, every crossed
+// level must have an explicit row so a malformed runtime PVF cannot silently
+// underpay the ledger.
+func (t *Tables) TotalTechniquePoints(level int) (int, error) {
+	if t == nil || level <= 0 {
+		return 0, fmt.Errorf("%w: TP level=%d", ErrLevelOutOfRange, level)
+	}
+	firstLevel := t.firstTechniquePointLevel()
+	if firstLevel == 0 || level < firstLevel {
+		return 0, nil
+	}
+	total := 0
+	for current := firstLevel; current <= level; current++ {
+		award, ok := t.tpByLevel[current]
+		if !ok {
+			return 0, fmt.Errorf("%w: %s [tp table] missing level=%d", ErrTableMalformed, SkillPointTablePath, current)
+		}
+		if award > math.MaxInt-total {
+			return 0, fmt.Errorf("%w: %s cumulative TP overflow at level=%d", ErrTableMalformed, SkillPointTablePath, current)
+		}
+		total += award
+	}
+	return total, nil
+}
+
+func (t *Tables) firstTechniquePointLevel() int {
+	first := 0
+	if t == nil {
+		return first
+	}
+	for level := range t.tpByLevel {
+		if first == 0 || level < first {
+			first = level
+		}
+	}
+	return first
+}
+
+// AdvanceSkillPoints adds only the PVF awards for levels crossed since the
+// ledger's SyncedLevel. Both total and remaining SP grow by the same delta, so
+// previously spent SP is preserved exactly.
+func (t *Tables) AdvanceSkillPoints(points dnfrepo.SkillPointState, newLevel int) (SkillPointAdvance, error) {
+	result := SkillPointAdvance{Previous: points, New: points}
+	if points.SyncedLevel <= 0 || newLevel < points.SyncedLevel ||
+		points.TotalSP < 0 || points.RemainingSP < 0 || points.RemainingSP > points.TotalSP ||
+		points.TotalTP < 0 || points.RemainingTP < 0 || points.RemainingTP > points.TotalTP {
+		return SkillPointAdvance{}, fmt.Errorf("%w: target_level=%d ledger=%+v", ErrSkillPointLedger, newLevel, points)
+	}
+	spDelta := 0
+	tpDelta := 0
+	firstTPLevel := t.firstTechniquePointLevel()
+	for level := points.SyncedLevel + 1; level <= newLevel; level++ {
+		award, ok := t.SkillPointsAtLevel(level)
+		if !ok {
+			return SkillPointAdvance{}, fmt.Errorf("%w: %s missing level=%d", ErrTableMalformed, SkillPointTablePath, level)
+		}
+		if award > math.MaxInt-spDelta {
+			return SkillPointAdvance{}, fmt.Errorf("%w: %s SP delta overflow at level=%d", ErrTableMalformed, SkillPointTablePath, level)
+		}
+		spDelta += award
+		if firstTPLevel != 0 && level >= firstTPLevel {
+			tpAward, ok := t.TechniquePointsAtLevel(level)
+			if !ok {
+				return SkillPointAdvance{}, fmt.Errorf("%w: %s [tp table] missing level=%d", ErrTableMalformed, SkillPointTablePath, level)
+			}
+			if tpAward > math.MaxInt-tpDelta {
+				return SkillPointAdvance{}, fmt.Errorf("%w: %s TP delta overflow at level=%d", ErrTableMalformed, SkillPointTablePath, level)
+			}
+			tpDelta += tpAward
+		}
+	}
+	if spDelta > math.MaxInt-points.TotalSP || spDelta > math.MaxInt-points.RemainingSP ||
+		tpDelta > math.MaxInt-points.TotalTP || tpDelta > math.MaxInt-points.RemainingTP {
+		return SkillPointAdvance{}, fmt.Errorf("%w: SP/TP ledger overflow sp_delta=%d tp_delta=%d ledger=%+v", ErrSkillPointLedger, spDelta, tpDelta, points)
+	}
+	result.SPGain = spDelta
+	result.TPGain = tpDelta
+	result.New.TotalSP += spDelta
+	result.New.RemainingSP += spDelta
+	result.New.TotalTP += tpDelta
+	result.New.RemainingTP += tpDelta
+	result.New.SyncedLevel = newLevel
+	return result, nil
+}