Armor Perk Stack 0.26.0护甲六合一 0.26.0

中文版｜Armor Perk Stack 0.26.0​

护甲词条重配 · 五个可选项​

依赖：Bingus Shared Loader v16。模组需要加载器提供 API 1。​

推荐从 V4 精锐潜兵开始。​

V0、V1、V2、V3 会增加修饰符行数，仍有描述异常、引擎兼容或闪退风险。V4 保持原生行数，避免这类结构扩展，但不保证所有游戏版本和使用场景都稳定。​

模组改了什么​

为 RS-89、O-3、BFM-16 重新搭配被动效果，保留有用的原生词条，按档位加入抗性、换弹、护甲或人机工程加成。​

0.26.0 新增独立选项 V0 不死族，修改民主护佑被动，可与主档一起使用。​

选项​	定位​	底层行数​
V0 不死族	独立附加项，强化民主护佑	3 行 mod，无 stat
V1 不拉屎将军	O-3 单甲实验档	10 行 mod＋2 行 stat
V2 不拉屎亲卫	人机工程扩展档	RS-89／O-3／BFM-16 分别为 5／6／5 行 mod，O-3 另有 2 行 stat
V3 超级地球禁军	四行效果档	每件 4 行 mod，O-3 另有 2 行 stat
V4 精锐潜兵	原生行数、原位写回	三件分别为 2／3／3 行 mod，不新增 stat

V0 可以单开，也可以与任一主档同开。V1 至 V4 只选一个；误勾多个时，入口只加载优先级最高的主档：​

V4 ＞ V3 ＞ V2 ＞ V1​

推荐仅开启 V4，需要不死族时再加勾 V0。本次同步到 Arsenal 的配置也按此设置；重新导入 ZIP 后，请核对选项，不要假定导入器会保留勾选状态。​

文中的“行数”指底层修饰符行数，不等同于详情页的文字条数。“四防 75%”指火、毒、酸、电四类伤害减免 75%。​

各档配置​

V0 不死族｜民主护佑独立选项​

保留民主护佑的原生胸腔止血，再加入 RE 系列的肢体伤残防护，合计三行：​

· 免死概率设为 100%：民主护佑对应的倍率字段由 1.5 改为 2.0。​

· 保留胸腔流血免疫：原生效果不删、不换。​

· 避免严重肢体受伤的概率设为 100%：借用 RE 强化肩章的对应效果，倍率由 1.5 改为 2.0，目标是避免胳膊、腿等部位的伤残状态。​

只借肢体防护这一行，不复制 RE 的换弹、近战或隐藏 stat，也不直接修改 RE 的原记录。​

**这项配置不锁血、不免疫所有伤害，也不保证绕过所有死亡机制。**两项概率按原生字段的含义设置，实际游戏效果仍待实机确认。​

作用范围按共享被动判断。本地护甲数据包含 DP-00、DP-11、DP-40、DP-53，以及 B-22、TR-9。DP-8 使用其他被动，不在 V0 的修改范围。​

原生两行扩为三行，增加 16 字节的修饰符内容。新增数组放在模组自己的缓冲区，通过修改指针和条数引用，不是直接把这 16 字节写进邻居。游戏引擎能否始终正确处理扩展后的记录，仍需验证；三行较少，不代表没有风险。​

V0 与主档使用不同的目标被动、实现模块、状态文件和恢复记录，共存逻辑已通过离线测试。​

V1 不拉屎将军｜O-3 单甲实验档​

目标已经改为 O-3 自由之魂，不再是 RS-89。​

· 保留原生步行与奔跑加速，移除滑铲效果。​

· 主武器与支援武器换弹速度各增加 30%，连同所需的两行隐藏 stat 一起加入。​

· 武器人机工程增加 20，护甲增加 50。​

· 四防 75%、爆炸伤害抗性 50%、抗击倒免疫。​

· 敌人侦测范围减少 50%。​

共 10 行 mod＋2 行 stat。这是结构扩展幅度最大的主档，适合实验和对比。​

配方与 0.25.1 保持一致；0.26.0 将 V1 纳入共用运行逻辑，补齐状态更新、原始字节备份、副本维护和失败回滚。​

V2 不拉屎亲卫｜人机工程扩展档​

· RS-89 暗影典范：移动噪音减少 50%、敌人侦测范围减少 50%、爆炸伤害抗性 50%、护甲增加 50、武器人机工程增加 20。共 5 行 mod。​

· O-3 自由之魂：奔跑速度增加 15%、四防 75%、主武器与支援武器换弹速度各增加 30%、护甲增加 50、武器人机工程增加 20。共 6 行 mod＋2 行 stat。​

· BFM-16 坦克手：抗击倒免疫、爆炸伤害抗性 50%、护甲增加 50、投掷物初始携带量与持有上限增加 2、武器人机工程增加 20。共 5 行 mod。​

这是效果较多的扩展档，仍需实机验证。不同护甲的记录结构不同，不能把某件甲的“六行崩溃”直接当作所有护甲的统一上限，也不能据此给各档排出确定的风险名次。​

V3 超级地球禁军｜四行效果档​

· RS-89 暗影典范：移动噪音减少 50%、敌人侦测范围减少 50%、爆炸伤害抗性 50%、护甲增加 50。​

· O-3 自由之魂：奔跑速度增加 15%、四防 75%、主武器与支援武器换弹速度各增加 30%。步行加速与滑铲让位给本档效果。​

· BFM-16 坦克手：抗击倒免疫、爆炸伤害抗性 50%、护甲增加 50、投掷物初始携带量与持有上限增加 2。​

每件 4 行 mod，O-3 另带换弹所需的 2 行隐藏 stat。四行是本档的配置选择，不是经过证明的安全界限。​

V4 精锐潜兵｜推荐默认主档​

保持三件护甲的原生行数，只在原位置替换效果行或修改数值，不修改数组指针与条数，也不分配扩展缓冲区。​

· RS-89 暗影典范：移动噪音减少 75%、敌人侦测范围减少 50%。保留两行，不追加爆炸抗性或护甲。​

· O-3 自由之魂：奔跑速度增加 15%；步行加速行换成四防 75%；滑铲行换成护甲增加 50。共三行，不加入换弹或隐藏 stat。​

· BFM-16 坦克手：抗击倒免疫、护甲增加 50，雷达扫描行换成爆炸伤害抗性 50%。共三行，不加入投掷物加成。​

V4 避免了新增行带来的结构扩展风险。它仍会改写运行时效果，游戏更新、其他模组冲突或未覆盖的引擎行为都需要实际排查。​

共享被动与未修改的护甲​

本模组修改的是共享被动定义，效果不只限于列出的型号。​

例如 RS-100 Sanctioner、RS-67 Null Cipher 与 RS-89 共享相关被动，也会受到对应档位影响；O-3 和 BFM-16 的同被动护甲同理。​

护甲加成叠加在每件护甲自己的基础值上，不把所有型号统一设成某个数。对这里的 O-3、BFM-16 配置，增加 50 对应 50 → 100、75 → 125；其他重量级的结果按各自基础值计算。​

RE-1861 阅兵指挥官与 UF-16 Inspector 不作为直接改写目标。过去的扩展配置出现过兼容问题，因此保留这两类原记录。V0 使用的是 RE 肢体防护行的副本，来源记录不变；这不能进一步保证扩展配置不会引发其他引擎异常。​

运行方式与本版修复​

安装时由管理器部署补丁；运行后，模组在游戏进程内定位并修改被动定义，不改写游戏原始数据文件。日志和原始字节备份会写入本地磁盘。​

V0 至 V3 将重组后的数组放入自有缓冲区，再修改目标指针与条数；V4 直接在原位置写回效果行。退出时尝试恢复并核对，恢复不完整会报错。​

0.26.0 主要修复：​

· 时间片用完后继续扫描当前区域，不再漏掉后半段。​

· 一份副本失效时，保留其他副本的恢复记录。​

· 写入前备份并登记恢复信息；失败、抛错或部分写入时回滚，再回读验证。​

· 用独立低预算扫描寻找远处新副本，不只盯着已有副本附近。​

· 稳定期改为每两秒复查一次；状态按不同被动的覆盖数判断，不再把副本总数当作完整生效。​

扫描约在启动 60 秒后开始，初始扫描使用 64KB 块，每次预算最多 4ms 和 1MB。近邻检查约每 30 秒触发；远处发现扫描采用 1ms、256KB 的单次预算，完成一遍后间隔十分钟。​

单次系统读取无法中途打断，这些预算不构成硬实时保证。​

离线模拟中，三条补丁的四秒复查窗口从 2255 次读取降到 34 次。这是读取次数对比，不是实测帧率提升。​

安装与首次确认​

1. 安装并启用 Bingus Shared Loader v16，按加载器说明设置 Arsenal 优先级。​

2. 将 Armor Perk Stack 0.26.0 的目录或 ZIP 导入 HD2Arsenal。​

3. 主档建议只选 V4；需要不死族时额外勾 V0，也可只开 V0。​

4. 点击 Deploy，再启动或重启游戏。切换档位前先退出游戏，改完选项重新 Deploy。​

5. 等启动扫描开始并完成定位后，重新装备目标护甲，必要时出击一次，再查看状态和实际效果。​

**导入到库不等于部署成功。**重新导入 ZIP 后要检查勾选项。0.26.0 含公共入口和各档私有模块，不建议沿用旧版“任选目录、直接覆盖同名文件”的手动混装方式。​

日志目录：​

text复制​

%LOCALAPPDATA%/CowboyBingus/Helldivers2/Logs/​

项目​	主档 V1 至 V4​	独立选项 V0​
状态文件	ArmorPerkStack.STATUS.txt	ArmorPerkUndead.STATUS.txt
运行日志	ArmorPerkStack.log	ArmorPerkUndead.log
原始字节备份	ArmorPerkStack.orig.txt	ArmorPerkUndead.orig.txt

V1 正常覆盖 一类被动，V2 至 V4 各覆盖 三类，V0 覆盖 一类。请核对当前版本号和当前启动产生的记录。​

第一行结论在 .STATUS.txt 中，不要把旧日志当作本次状态。​

OK 表示找到目标并完成内存回读，不能代替重新装备，也不能证明所有实际效果都已验证。​

验证与使用限制​

0.26.0 已通过：​

· 196 项离线回归检查。​

· 4 项缺陷变异检测。​

· 12 项编译与原生接口检查。​

· 72 项独立发布验收，其中包含从库内实际归档重跑全部 31 种非空选项组合。​

这些检查覆盖了加载选择、模拟内存写入、故障回滚、多副本与恢复。原生接口测试只操作测试进程自己分配的内存。​

尚未完成本版的游戏内行为与联机验证，不能把离线通过写成实机稳定。​

模组会改变强度平衡。请在单人、固定队伍或其他参与者知情同意的情况下使用。即使修改发生在本机进程，也不应承诺“队友看不到”或“联机不影响别人”；同步表现与反作弊风险均不作保证。​

卸载与排错​

卸载前退出游戏，在 Arsenal 中停用模组，再 Deploy 并重新启动。手动安装过旧版的，请先确认自己添加的文件及编号，不要直接覆盖或随意删除不明补丁。​

没有效果时，依次检查：​

· 是否导入 0.26.0，选项是否正确，改完是否 Deploy 并重启。​

· 当前状态文件是否显示本次启动、正确版本及目标被动覆盖情况。​

· 定位完成后是否重新装备，必要时是否出击过一次。​

· 若启用 V0，是否查看了独立的 ArmorPerkUndead 状态文件。​

出现 FAILED、详情页异常或闪退时，先停用本模组并 Deploy，确认能否复现。保留所选档位、游戏版本、状态文件和对应日志再反馈。​

不要把首次闪退当作正常步骤，也不要因为使用 V4 就排除本模组。​

版本记录​

版本​	主要变化​
0.16.0	首个公开版本。
0.18.0	避开启动窗口，加入有限次数的定位尝试。
0.19.x	适配当时的新构建，核对 O-3 被动，加入删行、改值与多版本接管逻辑，并尝试原生行数方案。
0.20.0	回到四行效果配置，移除 RE-1861 直接修改目标。
0.21.x	收紧扫描开销，调整效果，并补上 O-3 换弹所需的隐藏 stat。
0.22.0	引入严格原生行数、原位写回的配置。
0.23.0	将扩展效果档与原生行数档分开提供。
0.24.0	重排为 V1 至 V4，新增人机工程扩展档，V4 作为推荐默认主档。
0.25.0	加入四防 75%；V2 至 V4 的 O-3 步行行改作四防，保留奔跑加速。
0.25.1	V1 从 RS-89 改为 O-3，去掉滑铲，形成 10 行 mod＋2 行 stat。
0.26.0	新增独立 V0 不死族；统一五档运行逻辑，修复续扫、副本管理与写入失败回滚，并更新入口、打包、说明和验证。

制作署名​

本模组由 AI 辅助制作：QWEN3.8MAX、GLM5.3MAX、step5.0preview。​

本版最终审核与收尾：GPT6astra。​

English｜Armor Perk Stack 0.26.0​

Armor Passive Reconfiguration · Five Selectable Options​

Requires Bingus Shared Loader v16. The mod requires Loader API 1.​

Start with V4 Elite Helldiver.​

V0, V1, V2, and V3 add modifier rows and still carry risks of broken descriptions, engine incompatibility, or crashes. V4 preserves native row counts and avoids that form of structural expansion, but it is not a stability guarantee for every game version or configuration.​

What This Mod Changes​

Armor Perk Stack reconfigures the passives used by RS-89, O-3, and BFM-16. It retains useful native effects and adds resistance, reload, armor, or ergonomics bonuses according to the selected tier.​

Version 0.26.0 adds V0 Undead, an independent option for Democracy Protects that can run alongside a base tier.​

Option​	Purpose​	Underlying Row Counts​
V0 Undead / 不死族	Independent Democracy Protects enhancement	3 modifier rows, no stat rows
V1 No-Poop General / 不拉屎将军	Experimental O-3-only configuration	10 modifier rows + 2 stat rows
V2 No-Poop Guard / 不拉屎亲卫	Expanded configuration with ergonomics bonuses	5/6/5 modifier rows for RS-89/O-3/BFM-16; O-3 also has 2 stat rows
V3 Super Earth Guard / 超级地球禁军	Four-row configuration	4 modifier rows per target; O-3 also has 2 stat rows
V4 Elite Helldiver / 精锐潜兵	Native row counts with in-place writes	2/3/3 modifier rows; no added stat rows

V0 works alone or alongside any one base tier. Select only one of V1–V4. If multiple base tiers are selected, the entrypoint loads only the highest-priority tier:​

V4 > V3 > V2 > V1​

The recommended selection is V4 only, with V0 added when wanted. The current Arsenal library synchronization uses that selection. Check the options when importing the ZIP again; do not assume the importer preserves their enabled states.​

“Row counts” refer to underlying modifier records, not the number of text lines shown on the armor details screen. “Four-element resistance 75%” means 75% damage reduction against fire, toxic, acid, and electrical damage.​

Option Details​

V0 Undead｜Independent Democracy Protects Option​

Preserves native chest-bleeding protection and adds the RE series’ limb-injury protection, for three modifier rows:​

· Sets the death-avoidance chance to 100%: changes Democracy Protects’ corresponding multiplier from 1.5 to 2.0.​

· Preserves native chest-bleeding immunity: the original effect is neither removed nor replaced.​

· Sets the chance to avoid grievous limb injuries to 100%: copies the corresponding effect from Reinforced Epaulettes and changes its multiplier from 1.5 to 2.0, targeting injury states affecting arms, legs, and similar body parts.​

Only the limb-protection row is copied. RE reload, melee, and hidden stat rows are not included, and the original RE record is not directly modified.​

This does not lock health, prevent all damage, or guarantee protection against every death mechanism. Both probabilities are configured according to the native fields’ meaning; their actual behavior still requires in-game verification.​

Coverage follows the shared passive. Local armor data includes DP-00, DP-11, DP-40, DP-53, B-22, and TR-9. DP-8 uses a different passive and is not targeted by V0.​

V0 expands the native two-row layout to three rows, adding 16 bytes of modifier content. The expanded array is stored in the mod’s own buffer and referenced through updated pointers and counts. Those additional 16 bytes are not written directly into the neighboring record. Whether the engine consistently handles the expanded record correctly remains unverified; fewer rows do not prove safety.​

V0 uses separate target passives, implementation modules, status files, and recovery records from the base tiers. Their coexistence logic has passed offline testing.​

V1 No-Poop General｜Experimental O-3 Configuration​

V1 now targets O-3 Free Spirit, not RS-89.​

· Preserves native walking and running bonuses; removes sliding.​

· Adds 30% reload speed for primary and support weapons, together with the two required hidden stat rows.​

· Adds 20 weapon ergonomics and 50 armor.​

· Provides four-element resistance 75%, explosive damage resistance 50%, and knockdown immunity.​

· Reduces enemy detection range by 50%.​

The configuration contains 10 modifier rows and 2 stat rows. It has the largest structural expansion among the base tiers and is intended for experimentation and comparison.​

Its effect recipe is unchanged from 0.25.1. Version 0.26.0 moves V1 onto the shared runtime and adds current status reporting, original-byte backups, copy maintenance, and failed-write rollback.​

V2 No-Poop Guard｜Expanded Ergonomics Configuration​

· RS-89: movement noise −50%, enemy detection range −50%, explosive damage resistance 50%, armor +50, and weapon ergonomics +20. 5 modifier rows.​

· O-3 Free Spirit: running speed +15%, four-element resistance 75%, primary and support weapon reload speed +30% each, armor +50, and weapon ergonomics +20. 6 modifier rows + 2 stat rows.​

· BFM-16: knockdown immunity, explosive damage resistance 50%, armor +50, starting throwable count and maximum capacity +2, and weapon ergonomics +20. 5 modifier rows.​

This is an effect-heavy expanded configuration and still requires in-game verification. Armor records have different layouts. A six-row crash observed on one armor does not establish a universal limit, nor does it justify a definitive risk ranking across tiers.​

V3 Super Earth Guard｜Four-Row Configuration​

· RS-89: movement noise −50%, enemy detection range −50%, explosive damage resistance 50%, and armor +50.​

· O-3 Free Spirit: running speed +15%, four-element resistance 75%, and primary and support weapon reload speed +30% each. The native walking and sliding effects are replaced.​

· BFM-16: knockdown immunity, explosive damage resistance 50%, armor +50, and starting throwable count and maximum capacity +2.​

Each target uses 4 modifier rows. O-3 also receives the 2 hidden stat rows required for the reload bonuses. Four rows are a configuration choice, not a proven safety threshold.​

V4 Elite Helldiver｜Recommended Base Tier​

Preserves native row counts and replaces effect rows or values in place. Array pointers and counts remain unchanged, and no expansion buffer is allocated.​

· RS-89: movement noise −75% and enemy detection range −50%. Keeps two rows, without adding explosive resistance or armor.​

· O-3 Free Spirit: running speed +15%; the walking row becomes four-element resistance 75%; the sliding row becomes armor +50. Three rows, with no reload bonuses or added hidden stats.​

· BFM-16: knockdown immunity and armor +50; the radar-scanning row becomes explosive damage resistance 50%. Three rows, without the throwable bonus.​

V4 avoids the structural expansion caused by additional rows. It still changes runtime effects, so game updates, conflicts with other mods, and untested engine behavior must be investigated when problems occur.​

Shared Passives and Untargeted Armor​

This mod changes shared passive definitions, so its effects are not limited to the listed armor models.​

For example, RS-100 Sanctioner and RS-67 Null Cipher share the relevant passive with RS-89 and receive the corresponding changes. The same principle applies to armor sharing O-3’s or BFM-16’s passives.​

Armor bonuses are added to each armor’s own base rating rather than setting every model to a fixed number. For the O-3 and BFM-16 configurations described here, +50 corresponds to 50 → 100 and 75 → 125. Other weight classes use their own base ratings.​

RE-1861 and UF-16 Inspector are not direct patch targets. Earlier expanded configurations encountered compatibility problems, so their original records are retained. V0 uses a copy of the RE limb-protection row without altering its source record. That does not guarantee that expanded configurations cannot cause other engine-side problems.​

Runtime Behavior and Fixes​

The mod manager deploys the patch during installation. At runtime, the mod locates and changes passive definitions in the game process without rewriting the game’s original data files. Logs and original-byte backups are written to local disk.​

V0–V3 place their rebuilt arrays in separately allocated buffers and update the target pointers and counts. V4 writes effect rows back in place. On shutdown, the mod attempts restoration and verifies the result; incomplete restoration is reported as a failure.​

Key fixes in 0.26.0:​

· Resumes the unfinished portion of a memory region after a time slice instead of skipping it.​

· Preserves recovery records for surviving copies when another copy becomes invalid.​

· Saves original bytes and registers recovery state before target writes; failures, exceptions, and partial writes trigger rollback with readback verification.​

· Uses a separate low-budget scan to discover distant new copies rather than watching only nearby memory.​

· Rechecks stable patches every two seconds and determines completion by distinct passive coverage, not total copy count.​

Scanning begins approximately 60 seconds after startup. Initial scans use 64KB chunks with a per-step budget of 4ms and 1MB. Nearby checks are triggered approximately every 30 seconds. Distant-copy discovery uses a 1ms/256KB per-step budget and waits ten minutes after completing a pass.​

Individual system reads cannot be interrupted mid-call, so these budgets are not hard real-time guarantees.​

In the offline simulation, a four-second verification window covering three patches dropped from 2,255 reads to 34. This measures read counts, not an observed FPS improvement.​

Installation and First Check​

1. Install and enable Bingus Shared Loader v16, following the loader’s instructions for Arsenal priority.​

2. Import the Armor Perk Stack 0.26.0 directory or ZIP into HD2Arsenal.​

3. Select V4 as the recommended base tier. Add V0 for Undead, or enable V0 by itself.​

4. Click Deploy, then start or restart the game. Exit the game before switching tiers, and Deploy again after changing options.​

5. Once startup scanning has located the targets, re-equip the relevant armor. Enter a mission if necessary, then check both status and actual behavior.​

Importing a mod into the library does not deploy it. Check the enabled options after importing a ZIP. Version 0.26.0 includes a shared entrypoint and private implementation modules; the old approach of manually mixing option folders by overwriting identically named files is not recommended.​

Log directory:​

text复制​

%LOCALAPPDATA%/CowboyBingus/Helldivers2/Logs/​

File Type​	Base Tiers V1–V4​	Independent V0 Option​
Status	ArmorPerkStack.STATUS.txt	ArmorPerkUndead.STATUS.txt
Runtime log	ArmorPerkStack.log	ArmorPerkUndead.log
Original-byte backup	ArmorPerkStack.orig.txt	ArmorPerkUndead.orig.txt

V1 normally covers one passive, V2–V4 cover three each, and V0 covers one. Check the version and confirm that the records belong to the current launch.​

The first-line status summary is in .STATUS.txt. Do not mistake an old log for the current state.​

OK means the targets were located and memory readback succeeded. It does not replace re-equipping the armor or prove that every gameplay effect has been verified.​

Validation and Usage Limits​

Version 0.26.0 has passed:​

· 196 offline regression checks.​

· 4 mutation tests.​

· 12 compilation and native-interface checks.​

· 72 independent release checks, including rerunning all 31 nonempty option combinations from the actual library archives.​

These checks cover option selection, simulated memory writes, failure rollback, multiple copies, and restoration. Native-interface tests operate only on memory allocated by the test process itself.​

In-game behavior and multiplayer validation for this release are still pending. Offline success must not be presented as proven in-game stability.​

The mod changes gameplay balance. Use it in solo play, with a regular group, or with the informed consent of other participants. Changes occurring in the local process do not justify promises that teammates cannot see them or that other players are unaffected. Multiplayer synchronization and anti-cheat safety are not guaranteed.​

Removal and Troubleshooting​

Exit the game, disable the mod in Arsenal, Deploy, and restart. If an older version was installed manually, identify the files and patch numbers you added before changing anything. Do not overwrite or delete unidentified patches.​

If the mod appears inactive, check the following:​

· Is version 0.26.0 imported, are the intended options selected, and did you Deploy and restart?​

· Does the current status file show the correct version, current launch, and expected passive coverage?​

· Did you re-equip after target discovery, or enter a mission if necessary?​

· If V0 is enabled, did you check its separate ArmorPerkUndead status file?​

For FAILED states, broken armor descriptions, or crashes, disable the mod and Deploy before checking whether the issue still occurs. Include the selected options, game version, status files, and corresponding logs in a report.​

Do not treat a first-launch crash as a normal installation step, and do not rule out this mod merely because V4 is selected.​

Version History​

Version​	Main Changes​
0.16.0	First public release.
0.18.0	Delayed scanning beyond startup and introduced a limited number of discovery attempts.
0.19.x	Adapted to the game builds available at the time, corrected O-3 passive identification, added row removal, value editing, and version takeover logic, and experimented with native row counts.
0.20.0	Returned to four-row effect configurations and removed RE-1861 as a direct patch target.
0.21.x	Reduced scanning overhead, adjusted effects, and added the hidden stats required for O-3 reload bonuses.
0.22.0	Introduced strict native row counts with in-place writes.
0.23.0	Offered expanded-effect and native-row configurations separately.
0.24.0	Reorganized the options into V1–V4, added the expanded ergonomics tier, and made V4 the recommended base tier.
0.25.0	Added four-element resistance 75%; V2–V4 replaced O-3’s walking row with resistance while retaining the running bonus.
0.25.1	Moved V1 from RS-89 to O-3, removed sliding, and established its 10-modifier/2-stat layout.
0.26.0	Added independent V0 Undead; unified the five options’ runtime, repaired scan continuation, copy management, and failed-write rollback, and updated the entrypoint, packaging, documentation, and validation.

Credits​

AI-assisted development: QWEN3.8MAX, GLM5.3MAX, and step5.0preview.​

Final review and finishing work for this release: GPT6astra.​

​