在《杀戮尖塔 2》那片被诅咒与救赎反复撕扯的尖塔之下，全新角色卡芙卡以 “持续伤害操控者” 的身份登场，她不依赖瞬间爆发的狂轰滥炸，而是用层层叠叠的毒素、灼烧与创伤，将敌人拖入慢性死亡的深渊，在回合的间隙里收割生命，在尖塔的阴影中编织属于自己的杀戮韵律。
一、角色设定：从阴影中走来的毒语者
卡芙卡曾是尖塔外围 “低语回廊” 的守夜人，她的家族世代以 “调和尖塔的腐蚀之力” 为使命，却在一场异变中被毒素反噬 —— 她的血液里流淌着尖塔最原始的腐蚀能量，指尖能轻易触碰到无形的 “伤害印记”，每一次触碰都能让印记生根、蔓延、爆发。
她的外形带着尖塔独有的诡谲美感：深紫与墨黑交织的长袍上，绣着不断蠕动的毒藤纹路；银白的长发束成高马尾，发梢沾着淡绿色的毒素结晶；眼眸是深邃的暗紫色，看向敌人时，仿佛能看穿对方的血肉，精准找到最脆弱的印记节点。她的武器不是刀剑，而是一对镶嵌着毒晶的指套，每一次挥出，都带着 “标记 - 蔓延 - 引爆” 的连锁回响。
卡芙卡的核心特质是 “掌控持续伤害（DOT）”—— 她不追求一回合清空敌人血条，而是通过卡牌与遗物的联动，让敌人身上的每一层 “中毒”“灼烧”“创伤” 都成为定时炸弹，在她的指令下提前引爆、反复叠加，让敌人在每一个回合都承受撕心裂肺的痛苦，直至彻底崩塌。
二、核心机制：印记与引爆的循环艺术
卡芙卡的战斗体系围绕 “挂印记→叠层数→引爆伤害→再挂印记” 的循环展开，所有卡牌与天赋都服务于这一核心逻辑，让持续伤害从 “被动等待” 变成 “主动操控”。
- 核心天赋：毒脉回响
每当敌人身上的 ** 持续伤害状态（中毒 / 灼烧 / 创伤）** 触发时，卡芙卡获得 1 点 “毒能”；
消耗 2 点毒能，可立即引爆目标身上所有持续伤害，造成等同于当前层数总和的额外伤害（不消耗持续伤害层数）；
回合结束时，若拥有≥3 点毒能，自动为所有敌人叠加 1 层 “腐蚀印记”（每回合造成 2 点真实伤害，可与其他持续伤害叠加）。
- 初始卡组：毒刺的开端
卡芙卡的初始卡组没有高额伤害的攻击牌，却满是 “标记” 与 “铺垫”，为后续的连锁爆发埋下伏笔：
毒刺（1 费攻击）：造成 5 点伤害，为目标叠加 1 层中毒；
蔓延（1 费技能）：抽 1 张牌，将目标身上的 1 层持续伤害复制给所有敌人；
腐蚀之触（1 费技能）：获得 3 点格挡，为自己叠加 1 层 “毒抗”（减少 1 点受到的持续伤害，最多叠加 5 层）；
回响（1 费攻击）：造成 3 点伤害，若目标有持续伤害，额外获得 1 点毒能。
- 特色卡牌体系：从铺垫到爆发
卡芙卡的卡牌分为三大类，完美支撑 “持续伤害操控” 的核心玩法：
（1）印记叠加类：铺好死亡的基石
剧毒注射（2 费攻击）：造成 8 点伤害，为目标叠加 3 层中毒，若目标已有中毒，额外叠加 2 层；
灼烧烙印（2 费技能）：为所有敌人叠加 2 层灼烧，每有 1 层灼烧，本回合获得 1 点力量；
创伤穿刺（3 费攻击）：造成 12 点伤害，为目标叠加 3 层创伤（每回合造成等同于层数的伤害，无法被格挡），并获得 2 点毒能。
（2）引爆与增幅类：收割的关键
毒爆（2 费技能）：消耗所有毒能，每点毒能引爆目标身上 1 层持续伤害，造成额外伤害；
连锁回响（3 费技能）：引爆所有敌人身上的持续伤害，每引爆 1 层，为所有敌人叠加 1 层新的持续伤害（随机为中毒 / 灼烧 / 创伤）；
毒脉增幅（1 费技能）：本回合内，所有持续伤害的触发伤害提升 50%，引爆伤害提升 100%。
（3）生存与续航类：在毒潮中站稳脚跟
毒盾（2 费技能）：获得等同于当前毒能 ×3 的格挡，每有 1 层自身毒抗，额外获得 2 点格挡；
汲取毒素（1 费技能）：移除自身 1 层毒抗，抽 2 张牌，获得 2 点毒能；
腐蚀转化（3 费技能）：将敌人身上所有持续伤害层数的 50%，转化为自身的格挡与毒能。
三、遗物联动：让毒潮席卷尖塔
卡芙卡的专属遗物进一步强化了持续伤害的操控性，让她的玩法更具策略深度：
毒晶核心（专属遗物）：初始获得 2 点毒能，每叠加 1 层持续伤害，额外获得 1 点毒能（每回合最多触发 5 次）；
回响项链：引爆持续伤害时，有 50% 概率不消耗毒能；
腐蚀指环：敌人身上的持续伤害层数≥5 时，每回合开始时自动引爆 1 层；
毒藤斗篷：受到伤害时，50% 概率为攻击者叠加 1 层中毒；
印记沙漏：所有持续伤害的持续回合 + 1，引爆伤害额外提升 30%。
四、玩法流派：两种极致的毒杀艺术
- 叠层引爆流（新手友好）
核心思路：快速叠加多层持续伤害，用 “毒爆”“连锁回响” 一次性引爆，打出高额爆发；
关键卡牌：剧毒注射、灼烧烙印、毒爆、连锁回响；
遗物搭配：毒晶核心、印记沙漏、腐蚀指环；
玩法节奏：前期用 “毒刺”“蔓延” 快速挂印记，中期叠层 + 攒毒能，后期一回合引爆清场。
- 持续蔓延流（进阶策略）
核心思路：不追求瞬间引爆，而是让持续伤害无限蔓延、反复叠加，让敌人每回合都承受巨额伤害；
关键卡牌：蔓延、创伤穿刺、毒脉增幅、腐蚀转化；
遗物搭配：回响项链、毒藤斗篷、汲取毒素；
玩法节奏：前期保证生存，中期用 “蔓延” 复制持续伤害，后期靠 “毒脉增幅” 提升伤害，靠 “腐蚀转化” 续航，让敌人在持续掉血中绝望。
五、角色魅力：在尖塔中独树一帜的杀戮美学
与《杀戮尖塔 2》中其他角色相比，卡芙卡的魅力在于 “反直觉的战斗逻辑”—— 她让玩家放弃 “速战速决” 的惯性，学会享受 “步步为营” 的策略快感。
她的战斗没有华丽的特效，却在每一层印记的叠加、每一次毒能的积攒、每一轮引爆的回响中，展现出极致的操控感。当敌人身上爬满中毒、灼烧、创伤的纹路，当每一次引爆都让血条暴跌，当毒能在指尖流转成死亡的韵律，玩家会真正感受到：卡芙卡的杀戮，不是粗暴的摧毁，而是一场精心编排的、属于持续伤害的艺术盛宴。
在尖塔的无尽轮回里，卡芙卡带着她的毒刺与回响而来，她不救赎，不怜悯，只以腐蚀之力，在阴影中写下属于自己的杀戮篇章 —— 每一个被她标记的敌人，都将在无尽的痛苦中，成为她毒脉里的一抹回响。
https://github.com/dakeai-gpt/-/issues/239
https://github.com/dakeai-gpt/-/issues/240
https://github.com/dakeai-gpt/-/issues/241
https://github.com/dakeai-gpt/-/issues/242
https://github.com/dakeai-gpt/-/issues/243
https://github.com/dakeai-gpt/-/issues/244
https://github.com/dakeai-gpt/-/issues/245
https://github.com/dakeai-gpt/-/issues/246
https://github.com/dakeai-gpt/-/issues/247
https://github.com/dakeai-gpt/-/issues/248
https://github.com/dakeai-gpt/-/issues/249
https://github.com/dakeai-gpt/-/issues/250
https://github.com/dakeai-gpt/-/issues/251
https://github.com/dakeai-gpt/-/issues/252
https://github.com/dakeai-gpt/-/issues/253
https://github.com/dakeai-gpt/-/issues/254
https://github.com/dakeai-gpt/-/issues/255
https://github.com/dakeai-gpt/-/issues/256
https://github.com/dakeai-gpt/-/issues/257
https://github.com/dakeai-gpt/-/issues/258
https://github.com/dakeai-gpt/-/issues/259
https://github.com/dakeai-gpt/-/issues/261
https://github.com/dakeai-gpt/-/issues/262
https://github.com/dakeai-gpt/-/issues/263
https://github.com/dakeai-gpt/-/issues/264
https://github.com/dakeai-gpt/-/issues/265
https://github.com/dakeai-gpt/-/issues/266
https://github.com/dakeai-gpt/-/issues/267
https://github.com/dakeai-gpt/-/issues/268
https://github.com/dakeai-gpt/-/issues/269
https://github.com/dakeai-gpt/-/issues/270
https://github.com/dakeai-gpt/-/issues/271
https://github.com/dakeai-gpt/-/issues/272
https://github.com/dakeai-gpt/-/issues/273
https://github.com/dakeai-gpt/-/issues/274
https://github.com/dakeai-gpt/-/issues/275
https://github.com/dakeai-gpt/-/issues/276
https://github.com/dakeai-gpt/-/issues/277
https://github.com/dakeai-gpt/-/issues/278
https://github.com/dakeai-gpt/-/issues/279
https://github.com/dakeai-gpt/-/issues/280
https://github.com/dakeai-gpt/-/issues/281
https://github.com/dakeai-gpt/-/issues/282
https://github.com/dakeai-gpt/-/issues/283
https://github.com/dakeai-gpt/-/issues/284
https://github.com/dakeai-gpt/-/issues/285
https://github.com/dakeai-gpt/-/issues/286
https://github.com/dakeai-gpt/-/issues/287
https://github.com/dakeai-gpt/-/issues/288
https://github.com/dakeai-gpt/-/issues/289
https://github.com/dakeai-gpt/-/issues/290
https://github.com/dakeai-gpt/-/issues/291
https://github.com/dakeai-gpt/-/issues/292
https://github.com/dakeai-gpt/-/issues/293
https://github.com/dakeai-gpt/-/issues/294
https://github.com/dakeai-gpt/-/issues/295
https://github.com/dakeai-gpt/-/issues/296
https://github.com/dakeai-gpt/-/issues/297
https://github.com/dakeai-gpt/-/issues/298
https://github.com/dakeai-gpt/-/issues/299
https://github.com/dakeai-gpt/-/issues/300
https://github.com/dakeai-gpt/-/issues/301
https://github.com/dakeai-gpt/-/issues/302
https://github.com/dakeai-gpt/-/issues/303
https://github.com/dakeai-gpt/-/issues/304
https://github.com/dakeai-gpt/-/issues/305
https://github.com/dakeai-gpt/-/issues/306
https://github.com/dakeai-gpt/-/issues/307
https://github.com/dakeai-gpt/-/issues/308
https://github.com/dakeai-gpt/-/issues/309
https://github.com/dakeai-gpt/-/issues/310
https://github.com/dakeai-gpt/-/issues/311
https://github.com/dakeai-gpt/-/issues/312
https://github.com/dakeai-gpt/-/issues/313
https://github.com/dakeai-gpt/-/issues/314
https://github.com/dakeai-gpt/-/issues/315
https://github.com/dakeai-gpt/-/issues/316
https://github.com/dakeai-gpt/-/issues/317
https://github.com/dakeai-gpt/-/issues/318
https://github.com/dakeai-gpt/-/issues/319
https://github.com/dakeai-gpt/-/issues/320
https://github.com/dakeai-gpt/-/issues/321
https://github.com/dakeai-gpt/-/issues/322
https://github.com/dakeai-gpt/-/issues/323
https://github.com/dakeai-gpt/-/issues/324
https://github.com/dakeai-gpt/-/issues/325
https://github.com/dakeai-gpt/-/issues/326
https://github.com/dakeai-gpt/-/issues/327
https://github.com/dakeai-gpt/-/issues/328
https://github.com/dakeai-gpt/-/issues/329
https://github.com/dakeai-gpt/-/issues/330
https://github.com/dakeai-gpt/-/issues/331
https://github.com/dakeai-gpt/-/issues/332
https://github.com/dakeai-gpt/-/issues/333
https://github.com/dakeai-gpt/-/issues/334
https://github.com/dakeai-gpt/-/issues/335
https://github.com/dakeai-gpt/-/issues/336
https://github.com/dakeai-gpt/-/issues/337
https://github.com/dakeai-gpt/-/issues/338
https://github.com/dakeai-gpt/-/issues/339
https://github.com/dakeai-gpt/-/issues/340
https://github.com/dakeai-gpt/-/issues/341
https://github.com/dakeai-gpt/-/issues/342
https://github.com/dakeai-gpt/-/issues/343
https://github.com/dakeai-gpt/-/issues/344
https://github.com/dakeai-gpt/-/issues/345
https://github.com/dakeai-gpt/-/issues/346
https://github.com/dakeai-gpt/-/issues/347
https://github.com/dakeai-gpt/-/issues/348
https://github.com/dakeai-gpt/-/issues/349
https://github.com/dakeai-gpt/-/issues/350
https://github.com/dakeai-gpt/-/issues/351
https://github.com/dakeai-gpt/-/issues/352
https://github.com/dakeai-gpt/-/issues/353
https://github.com/dakeai-gpt/-/issues/354
https://github.com/dakeai-gpt/-/issues/355
https://github.com/dakeai-gpt/-/issues/356
https://github.com/dakeai-gpt/-/issues/357
https://github.com/dakeai-gpt/-/issues/358
https://github.com/dakeai-gpt/-/issues/359
https://github.com/dakeai-gpt/-/issues/360
https://github.com/dakeai-gpt/-/issues/361
https://github.com/dakeai-gpt/-/issues/362
https://github.com/dakeai-gpt/-/issues/363
https://github.com/dakeai-gpt/-/issues/364
https://github.com/dakeai-gpt/-/issues/365
https://github.com/dakeai-gpt/-/issues/366
https://github.com/dakeai-gpt/-/issues/367
https://github.com/dakeai-gpt/-/issues/368
https://github.com/dakeai-gpt/-/issues/369
https://github.com/dakeai-gpt/-/issues/370
https://github.com/dakeai-gpt/-/issues/371
https://github.com/dakeai-gpt/-/issues/372
https://github.com/dakeai-gpt/-/issues/373
https://github.com/dakeai-gpt/-/issues/374
https://github.com/dakeai-gpt/-/issues/375
https://github.com/dakeai-gpt/-/issues/376
https://github.com/dakeai-gpt/-/issues/377
https://github.com/dakeai-gpt/-/issues/378
https://github.com/dakeai-gpt/-/issues/379
https://github.com/dakeai-gpt/-/issues/380
https://github.com/dakeai-gpt/-/issues/381
https://github.com/dakeai-gpt/-/issues/382
https://github.com/dakeai-gpt/-/issues/383
https://github.com/dakeai-gpt/-/issues/384
https://github.com/dakeai-gpt/-/issues/385
https://github.com/dakeai-gpt/-/issues/386
https://github.com/dakeai-gpt/-/issues/387
https://github.com/dakeai-gpt/-/issues/388
https://github.com/dakeai-gpt/-/issues/389
https://github.com/dakeai-gpt/-/issues/390
https://github.com/dakeai-gpt/-/issues/391
https://github.com/dakeai-gpt/-/issues/392
https://github.com/dakeai-gpt/-/issues/393
https://github.com/dakeai-gpt/-/issues/394
https://github.com/dakeai-gpt/-/issues/395
https://github.com/dakeai-gpt/-/issues/396
https://github.com/dakeai-gpt/-/issues/397
https://github.com/dakeai-gpt/-/issues/398
https://github.com/dakeai-gpt/-/issues/399
https://github.com/dakeai-gpt/-/issues/400
https://github.com/dakeai-gpt/-/issues/401
https://github.com/dakeai-gpt/-/issues/402
https://github.com/dakeai-gpt/-/issues/403
https://github.com/dakeai-gpt/-/issues/404
https://github.com/dakeai-gpt/-/issues/405
https://github.com/dakeai-gpt/-/issues/406
https://github.com/dakeai-gpt/-/issues/407
https://github.com/dakeai-gpt/-/issues/408
https://github.com/dakeai-gpt/-/issues/409
https://github.com/dakeai-gpt/-/issues/410
https://github.com/dakeai-gpt/-/issues/411
https://github.com/dakeai-gpt/-/issues/412
https://github.com/dakeai-gpt/-/issues/413
https://github.com/dakeai-gpt/-/issues/414
https://github.com/dakeai-gpt/-/issues/415
https://github.com/dakeai-gpt/-/issues/416
https://github.com/dakeai-gpt/-/issues/417
https://github.com/dakeai-gpt/-/issues/418
https://github.com/dakeai-gpt/-/issues/419
https://github.com/dakeai-gpt/-/issues/420
https://github.com/dakeai-gpt/-/issues/421
https://github.com/dakeai-gpt/-/issues/422
https://github.com/dakeai-gpt/-/issues/423
https://github.com/dakeai-gpt/-/issues/424
https://github.com/dakeai-gpt/-/issues/425
https://github.com/dakeai-gpt/-/issues/426
https://github.com/dakeai-gpt/-/issues/427
https://github.com/dakeai-gpt/-/issues/428
https://github.com/dakeai-gpt/-/issues/429
https://github.com/dakeai-gpt/-/issues/430
https://github.com/dakeai-gpt/-/issues/431
https://github.com/dakeai-gpt/-/issues/432
https://github.com/dakeai-gpt/-/issues/433
https://github.com/dakeai-gpt/-/issues/434