The owner's decision (2026-10-03, after community feedback): 深巡 S2 行动能力剥夺, 雷蛇 S2 反击电弧, 号角 S2 暴风号令
and S3 终极防线, 灰毫 S1 攻击力强化·γ型 and S2 专注轰击 take the basic strategy (DEFAULT) instead of the 下半 TANK class
row's TAKE_DAMAGE. Every other MANUAL 重装 skill, 深巡 S1 included, still waits for a hit.
- tools/build-data.mjs TRIGGER_DEVIATIONS: a small table keyed by the normal chess id and the skill id (the elite
follows through baseId), never by the skill id alone - 灰毫 S1 is the generic skcom_atk_up[3]. resolveTrigger applies
it last; rawRule keeps the official TAKE_DAMAGE; validateAll fails the build when an entry no longer meets a
TAKE_DAMAGE row. data/chess.json rebuilt with --offline: exactly the 12 records' rule changes.
- PR #12's kit lines (trigger 'DEFAULT' on 深巡 / 雷蛇 S2) stay and agree with the data; their comments and tests now
give this reason instead of the community summary.
- Tests: test/sim/feedback1-tank-triggers.test.js (the data, kit = data, a real battle on 战场#01(下半) in which each of
the six casts with an enemy in range before any hit and each of the other 46 MANUAL 重装 records casts on the first
hit); data.test.js trigger expectations; the 灰毫 / 号角 kit tests follow; docs-consistency.
- Docs: DESIGN §21.29 (new), §5.6, §19.2, §20.2, §21.20 (settled row), §21.23; DATA, SIM, PLAYING, CHANGELOG 0.1.1;
research 03 §1.4 drops the arknights.wiki.gg "offensive / defensive" summary and records the activity_table history
(上半 had no TANK row, 下半 added it) and this deviation; research 00-INDEX.
Conflicts resolved keeping both sides:
- bandDraft.js header: the 本局禁用 mark of a strategy tied to a switched-off bond, then the 本局信息 dialog; the code
merged cleanly (the card's 本局禁用 tag and the detail note sit beside the dialog's state and button).
- draft.css: the 本局禁用 tag / note rules, then the 本局信息 entry and dialog block (its header cites DESIGN §21.28).
- CHANGELOG 0.1.1 界面: the gaps and issue #5 / #8 lines, then the dialog line.
- PLAYING §6: the disabled-bond line says where to look again (draft dialog, in-game 🔍); the 变形同构体 line keeps its
pairing explanation.
- DESIGN: the branch's "§21.26 Match info in the strategy draft" becomes §21.28; its two §10 references follow.
Conflicts resolved keeping both sides:
- effectsMeta.js header: the QA residual's "every [holder, item] pair, taken before the first item runs" and "an item
moved during it runs at most once", with the cell branch's wording of what 突变细胞 does mid-walk.
- DESIGN §6.6: the band pick that skips a strategy tied to a switched-off bond and the deployment of the operator a
transformation gains / the bench shed that spares it.
- DESIGN §21.25: PR #2's bench rule credited (the cell branch) plus the QA residual's snapshot-timing sentence.
- PLAYING: the new 突变细胞 sentence with the residual's hammer sentence (revives, borrowed hammers, the lend window).
- docs-consistency: both new tests (batch 6 QA residuals; 突变细胞 after the WA merge).
The hammer lock per deployment for borrowers and in-place revives (both [ASSUMED] where no source decides), 卢西恩 / 锏
casting for an airborne 蒂比, rule 3 for a summon that is gone, the dispatcher snapshot taken before the first item;
PLAYING and CHANGELOG no longer say every debuff misses an airborne 蒂比 (ground-enemy auras and counters still reach
her); research 04's hammer line says once per deployment. docs-consistency pins it.
- DESIGN §21.1 states the corrected rule with its evidence: bilibili BV1vzyVBuEN9 (上半, about 8:24) and
BV1Qkw1zMEoR (下半, about 7:25) show the carrier's tile empty at the next prep, one more deployment left
and the new operator on the bench; the PRTS 备注 is a destroy followed by a gain; PRTS 帮助 sends what a
player gains to the 手牌区; the closed PR #2 said the same (its merge path not adopted). The WA
keep-the-tile rule and the carrier-tile merge exception are withdrawn; the bench carrier and the
hand-and-temp-full case stay [ASSUMED].
- §20.11 / §20.9 no longer count a transformed carrier's tile; §21.20 lists the rule as official;
§21.25 credits PR #2's bench rule; §6.2 and §6.6 follow; the §21 intro maps it.
- META (bot, ctx.transform, built-ins, merge, notes), PLAYING, DATA (items.json note), research 04
(addendum 2, implFormula, open question 6) and research 01 row 7 follow.
- One CHANGELOG 0.1.1 line (PR #2 指出).
- docs-consistency pins the docs against the code (no fromKey / returnItems left, the gain through
acquireChess).
DESIGN §21.26 describes the 变形同构体 pairing list, the harmony mark in
the bond views and the bot's band pick; the normative lines of §6.3, §6.6,
§8.2 / §8.3 and the §21.11 note point at it. META documents the band
pick and the harmony field, PLAYING explains both for players, and the
CHANGELOG 0.1.1 section gets two 界面 lines and one 人机 line.
DESIGN maps F1-F5 in the §21 intro, adds the hammer-per-deployment row (user 2026-10-03) to
§21.20 with the batch's official rules and [ASSUMED] items, and rewrites the normative lines:
§3 fenced tiles, §5.3 liftoff, §5.4 fatal / ignoreSelect / redeploy, §5.5 bodies, blocking and
enemy selection, §6.4 onBuy and the item snapshot, §8.2 down row/col, §18.3, §19.7. CHANGELOG
0.1.1, SIM, DATA and PLAYING follow; docs-consistency checks them against the code.
Merges the 7 WB commits after the first WB merge (d1f8a49..2cb84bd): R3 one of 10 fixed
sets (9 seen), R9 one of 6 boss groups picked by the matches they came in, R11 one of the
7 seen lists of 7; R11 families 14 / 4 / 4; the 机密商店 and the 战术决策 can offer the
same card twice (R11 shop: two VI, at least one V and one 盟约之币). data/choices.json
rebuilt with tools/build-data.mjs --offline (identical to the branch's). The tactic bot
test uses a real seat (the merged bot scores tactic cards on the player's bonds).
DESIGN §21.2 / §7 / §21.20 and CHANGELOG describe the 22-match model.
DATA: choices.json tacticDraft, the families header and cards.tactic rows.
META 1.2: 战术决策 after the 4 official drafts, identical cards separate.
PLAYING: the 机变 row says a 战术决策 can show the same card twice and
what R11 offers.
The 7 seen R11 lists stay picked uniformly, assumed because that invents no list. By the blocks of
effectChoiceInfoDict, R11 would fire bounty_hunter_8..15 (beside artifact_paid_4/5 and hardbuff_select),
so the block order is no longer cited as support for '1..7'. 7, 8 or 9 events fit the 14 drafts about
equally; with 8, an unseen list would come in about 1 R11 bounty draft in 8 (open question).
- one subsection per workstream (reports, verdict, cause, official rule and source, what changed, [ASSUMED], tests),
the integration follow-ups and the settled decisions (tactician rotation keeps in-range summons, user-confirmed)
- §2 §3 §5 §6 §7 §8 §9 §10 §13 §14 §16-§20 rewritten where the merged behaviour changed them
- docs-consistency checks §21 against the code; PLAYING notes the player-visible follow-ups
- R11 悬赏决策: the list is one of the 7 seen lists, uniform (slots 7, taken as the events
bounty_hunter_1..7, one block of effectChoiceInfoDict). No list is built from nothing any more;
the rule only completes the 4 lists seen with 6 cards, so 12_8 / 16_2 / 16_6 come in about
1 R11 bounty draft in 15 (14 drafts without one: about 39 %).
- R9: the groups hold 9, 9, 9, 9, 8 and 6 cards (comments, docs, test); match 4's base-only
group may be a draft of another group (open). 58 -> 59 bounty drafts in the build-data note.
- 机密商店: shopDraft.rounds [11]. The official composition (VI, VI, V, 盟约之币 + 2) applies at
R11 (绝境 / 终极, where the screenshots are); the earlier 标准 / 险境 shops keep the previous draw,
any tier I-VI per card with replacement (repeats still possible, about 26 %).
- Tests: R11 slots / block order / never-shown rate, R9 group sizes, the non-R11 shop draw.
F4 "干员深巡放在围栏围着的地块里不会主动攻击" is not reproduced: 深巡 (哨戒铁卫, range 2-2) attacks every
enemy on her range tiles from the fenced 围墙 tiles of every active stage, the Final Assault halves and
联防, through a real Match and in the browser's own battle. Locked by test/sim/feedback1f-fence.test.js.
Nearby rule fixed: PRTS 围墙 / 围栏 地形机制 "部署在其中的单位，若当前阻挡类型为'地面阻挡'则无法阻挡敌人".
A unit on a tile ground units cannot pass now blocks no ground enemy (Battle._blockerFor); before, an
enemy pushed or dragged to the fence edge (0.5 from the unit) was blocked there - 薄绿 S2 on a fenced
tile held the dragged enemies. Air blocking (blockFly against flyers) stays. Docs: SIM.md Blocking,
PLAYING.md tip on the fenced tiles.
Player report F5 after 0.1.0: "突袭盟约的干员会移动到濒死干员的尸体上".
Real. A knocked-out operator left the occupancy map at once and only its
home tile stayed reserved. A 突袭 member that fell after its jump lay on the
landing tile, which nothing reserved, so the next member's jump (same target,
same scoring) landed on it; and a member's timed redeploy came back on its home
where a teammate that had jumped there lay. Real matches (six real 突袭
members, 战场#07 / #08, R6 / R9 / R12, 96 battles): 6 placements on a body
(3 jumps, 3 home redeploys) before, 0 of 526 jumps after.
Official (PRTS 卫戍协议/帮助, 作战阶段 / 单位部署):
- "干员退场后…原地留下一个“倒地干员”…满足再部署条件时，移除场上的该倒地干员
并自动部署至该位置"
- "倒地干员所在地块视为可部署，但所有我方单位在此处的部署行为将被阻止"
- "若干员被击倒的位置为其他干员或召唤物的初始位置，则在被击倒后，尝试返回其
自身的初始位置"
Now (sim/Battle.js):
- A knocked-out operator lies on `unit.body`: the tile it fell on, or its
own home when it fell on another board piece's home and its home is free
(one attempt at the knock-out, assumed). x / y / tileR / tileC keep where
it fell, so the kill / death handlers still act there.
- Every redeploy comes back there (restTile): the timed one, 不屈's and
阿戈尔's. A retreated operator still comes back on its home.
- downOn(r, c): _deploy (redeploys, the 突袭 landing, summons), spawnDevice
and relocate refuse the tile for every ally of the field; isReservedTile
reports it, so every automatic picker skips it (突袭 raidTile, tactical
points, summon / device pickers, 维娜's lions). The kit helpers that kept
their own home-tile loop (tier1 summonTileFree, tier3 / tier4 freeTile) go
through isReservedTile.
- b.snap `down` entries gain the body tile: [id, respawnAt, respawnTime,
state, row, col]. DOWN_STATE.WAIT_TILE becomes a safeguard only.
Tests: test/sim/feedback1f-downed-tile.test.js (real Match / PlayerState /
BattleSpec runs with real 突袭 chess, the 突袭 jump, the engine refusals,
the return home); downelem and bonds_addon follow the official rule.
Review follow-up for E2 (not reproduced, no behaviour change):
- blockRun fills the skill's SP (and puts the patient back to 5 %) the
first tick the healer blocks, and tests 2 and 3 assert the skill is
cast or running while it blocks: all 56 healer loadouts now cast in
the window (half never did at the default initial SP).
- The reported 炎 boss round makes 录武官's skill ready when she first
holds 卢西恩 / 不祥幻影 and asserts it is cast or running then.
- The boss test says what it covers: 0 healer damage on all 20 boss /
hidden fields; healers block 卢西恩 / 不祥幻影 (h07_05) and a
patrolling 碎铳之簧 once its shield is broken (h08_02); the other
leaders and parts cannot be blocked by a block-1 unit.
- SIM.md: the healer qualifier follows the "one-line flip" sentence, so
"This replaces..." refers to the every-blocker rule again.
- ai.js header, Battle.blockedTargets and targeting.js: the blocked-first
rule is for attacks that hit enemies; a heal attack keeps selecting
injured allies (PRTS 卫戍协议/帮助).
- PLAYING.md: the healer exception names 咒愈师 as the one that hits.
The direction rules now rest on client data instead of the skill wording:
野鬃 S2's attack ability carries buff wildmn_s_2[force] of template
knockback[dir] (Knockback, _useSourceDirection true, force level -2 when
off direction); 琳琅诗怀雅 S3, 山 S3 and 莫斯提马 S3 use knockback[relative]
(radial); 圣聆初雪 S1 is KnockBackWithCharacterDirection (no correction).
Comments, SIM.md and the test header cite these. The generic kit header
notes its text keywords are only a fallback. PLAYING.md explains 野鬃's
push direction to players. No behaviour change.
- render/app.js: the views' unit info is built by the exported renderInfo(UnitInfo), which now keeps `form`;
addInfo dropped it, so a view built from fieldMeta (a teammate's field watched later, 联防 observers, a
reconnect) still drew the first form. Tested through renderInfo (fake PIXI) and the real enterBattle(meta)
path in headless Chrome (opt-in).
- content/enemies.js: rebirthCleanse at every 重生 (reborn, husk, statue): operator-applied buffs and
source-less statuses go, own talents / enemy-given buffs / field state stay (PRTS 特殊机制 §重生). A 逐火
knocked out while feared is no longer an unblockable ember.
- snapshot.js: an enemy's stealth bit only while its 隐匿 is on (not blocked, not revealed).
- Docs: area damage reaching an unblocked 隐匿 enemy is documented as an engine deviation (PRTS 作战机制
§AOE伤害判定), not as the rule.
- Every form change goes through content setForm(): the unit keeps it (e.form)
and UnitInfo publishes it, so a view built mid-battle from fieldMeta (a
watched teammate's field, 联防 observers, a reconnect) starts in the
current form instead of the first-form model.
- Enemy profile (dmgType) and Unit moving / form are initialised at spawn:
one object shape for the sim's hot loop (the boss-field bench is back at
the base speed).
- 转译基底·α: an HP loss cannot knock out its original form (no death clip).
- 假想敌：再生: 1 s 重生 and walking 傀儡 cited from PRTS; shields within 1.8
as the 再生状态 begins.
- Renderer: a 重生's closing clip (and the ember's stand-up clip) is timed
from the fx dur to end with it, so the next form attacks on its own clips.
- Docs: ASSETS FORMS list, SIM form fx, PLAYING 隐匿 wording.
Review follow-up to the water / tactical point change.
- Re-orienting 伺夜 / 缪尔赛思 in place while hand and temp are full could leave
a placed 狼群 / 流形 outside the new range (it had nowhere to go and stayed
put). The re-orientation is now refused with HAND_FULL when a summon it
would leave out has no stack and no free slot. On the other paths
(promotion, an owner moved with no room) such a summon leaves the board and
its stack comes back at the next round start (grantTokensFor), so no
out-of-range placement reaches the battle.
- Citation corrected: the 深水区 rule comes from the PRTS page 深水区 (地形信息
tile_deepsea: 部署类型 全部位, 地形机制 "拒绝部署（待补充）"), corroborated by
PRTS 作战机制 (地形标记 AdvancedBuildableMask lists 深水). There is no PRTS
page 地形.
- grid.js: the tie-break note now says what changed on 战场#08 (no gate
route; a few off-route smoothed steps of the boss field go around the water).
- Documented the known limit of the inactive act1 m05: its 特制水上平台 tiles
are deployable in the prep's deploy map only; the sim does not model the
canoes and keeps those tiles non-standable.
- gameLogic.summonRange reuses facing.js rangeTiles.
- tools/build-data.mjs bountyDraftHalf: exact faction-series counts (12 of 18 two-battle cards bring first-half
enemies, 5 outside the wave tables, enemyeffect_12_4 a second-half attachment), base-HP gradient, the user's
two-battle recall; counter-evidence (假想敌 named enemyInitial_*, the unread NGA R3 guide); the split marked
[ASSUMED] with its flip; data unchanged
- server/match/choices.js, docs/META.md, docs/DATA.md: the split marked [ASSUMED]
- m.private shop.rewardOffer.queued: offers waiting behind the shown one; the header says 之后还有 N 项 and the
reminder pill +N (a 定向投放 behind a promotion reward put off)
- tests: corrected headers, the stale bountyTiers fixture dropped, queued checks (unit + opt-in e2e on desktop
and 756x366)
- A pair Final Assault / Hidden Core field has one spawn list that both
players' handlers edit; each player's 0-2 swaps now come only from the
enemies whose route ends on that player's half (the half rule of the
field's bounties). Match passes the wave's routes and the player's side
('L' / 'R', null on a one-player field) in the boss / hidden
onBattleStart event; docs/META.md documents the two fields.
- The comment justifying elite eligibility was wrong (the leader rounds'
escorts are a mix, about two thirds elites, not all elites); reworded
with the measured numbers and kept [ASSUMED].
- The pair-field test now drives the real match path (Final Assault, then
the Hidden Core it unlocks) instead of dispatching by hand; a unit check
covers the half filter.
- render/spine.js: no attack (wind-up or 'atk') cuts a form's change clip short;
a knocked-out 逐火 no longer reports its last attack in the snapshot.
- test/render/feedback1-forms.browser.test.js (opt-in RENDER_E2E=1): the real
sim in headless Chrome on 战场#01 — 转译基底·α A_* → A_Die_B → B_* → B_Die,
a 深池逐火战士 knocked out in front of its blocker: Die → the 隐匿 ember on
Idle_2, beaten → Die_2.
- PLAYING.md: the 余烬 and 转译基底 rules for players.
Player reports #3 and #9 after the 0.1.0 release.
#3 On 战场#08(下半) 涨潮控制 (act2autochess_m04) operators could be placed in the
pool and 突袭 members jumped into it. The 深水区 (tile_deepsea) has buildableType
ALL in the level file, but it refuses deployment (PRTS 地形 深水区 "地形机制：拒绝
部署"; the season-1 战场#05 puts a 特制水上平台 - "在水上建立可以部署任意单位的
平台" - on every one of its water tiles).
- grid.js DEPLOY_REFUSED_TILES: the legend's effective build is NONE, so
canStand (突袭 landing tile, tactical points, summon tiles) never picks the
water; a hard-blocked 射击台 / mound tile takes no melee unit either.
- build-data: stages.json legend `buildable` is the effective type (the level's
value kept in `buildableType`), deployTiles exclude the water, a 特制水上平台
makes its tile deployable (act1 m05, inactive). board.js / gameLogic.js
deploy maps follow (water platforms included); bots plan on the same map.
- Gate routes are unchanged; the water now counts as non-blockable ground in
the path tie-break (audit lists updated).
#9 伺夜's 狼群 and 缪尔赛思's 流形 ("只能部署在召唤者攻击范围内", tokens.json
`ownerRange`) could be placed anywhere.
- PlayerState._legal / summonRange: a range-bound summon only on a tile of its
owner's attack range (loadout grid rotated by the owner's facing); swaps with
the owner are checked from the owner's new tile; re-orienting (or promoting)
the owner sends a summon left outside back onto its stack with a toast.
- The client lights and accepts the same tiles (gameLogic.summonRange), the bot
plans inside them, the invariants check it, the 伺夜 kit re-uses the piece's
tile only inside her range.
Tests: test/match/feedback1-placement, test/sim/feedback1-water, the opt-in
headless-Chrome test/ui/feedback1-placement.e2e (fastServer hook SP_STAGE);
existing summon / deploy / pathing tests moved off the water tiles.
sganggs