universal-dev-standards 5.17.0 → 6.0.0

package/bundled/skills/atdd-assistant/atdd-workflow.md

@@ -233,13 +233,12 @@ Convert acceptance criteria into executable, automatable tests.
 # File: features/password-reset.feature
 # Story: US-123 - Password Reset
 
-@US-123
 Feature: Password Reset
   As a registered user
   I want to reset my password via email
   So that I can regain access if I forget my password
 
-  @AC-1 @happy-path
+  @US-123 @AC-1 @happy-path
   Scenario: Request password reset
     Given I am on the login page
     And I have a registered account with email "user@example.com"
@@ -248,14 +247,14 @@ Feature: Password Reset
     Then I should see "Reset link sent to your email"
     And I should receive an email within 5 minutes
 
-  @AC-2 @error-handling
+  @US-123 @AC-2 @error-handling
   Scenario: Reset link expires after 24 hours
     Given I have requested a password reset
     When I click the reset link after 24 hours
     Then I should see "This link has expired"
     And I should be offered to request a new link
 
-  @AC-3 @security
+  @US-123 @AC-3 @security
   Scenario: Unregistered email shows generic message
     Given I am on the password reset page
     When I enter an unregistered email
@@ -525,13 +524,12 @@ Demonstrate the completed feature to stakeholders and get formal acceptance.
 ### Distilled Tests (Gherkin)
 
 ```gherkin
-@US-456
 Feature: Order Confirmation Email
 
   Background:
     Given the email service is configured
 
-  @AC-1
+  @US-456 @AC-1
   Scenario: Email sent on successful order
     Given I am logged in as "customer@example.com"
     And I have items in my cart
@@ -540,7 +538,7 @@ Feature: Order Confirmation Email
     And the email should be received within 5 minutes
     And the email should contain my order number
 
-  @AC-2
+  @US-456 @AC-2
   Scenario: Email contains complete order details
     Given I have completed an order with:
       | item       | quantity | price |
@@ -554,7 +552,7 @@ Feature: Order Confirmation Email
     And it should show my shipping address
     And it should show estimated delivery date
 
-  @AC-3
+  @US-456 @AC-3
   Scenario: Guest checkout receives email
     Given I am not logged in
     And I complete checkout as guest with email "guest@example.com"

package/bundled/skills/audit-assistant/SKILL.md

@@ -94,7 +94,7 @@ uds audit --report --dry-run
 
 ## Integration with Other Skills | 與其他技能的整合
 
-When suggested by other skills (e.g., `/checkin`, `/commit`, `/review`, `/sdd`), `/audit` can be invoked to:
+When suggested by other skills (e.g., `/checkin`, `/commit`, `/code-review`, `/sdd`), `/audit` can be invoked to:
 
 當被其他技能建議時，`/audit` 可根據情境執行針對性的檢查：
 
@@ -102,7 +102,7 @@ When suggested by other skills (e.g., `/checkin`, `/commit`, `/review`, `/sdd`),
 |---------|--------------------|------|
 | After `/checkin` failure | `/audit --health` | checkin 失敗後診斷 |
 | After `/commit` | `/audit --report` | 提交後回報發現 |
-| After `/review` finds friction | `/audit --friction` | 審查發現摩擦時 |
+| After `/code-review` finds friction | `/audit --friction` | 審查發現摩擦時 |
 | After `/sdd` spec creation | `/audit --patterns` | 規格建立後檢查覆蓋率 |
 
 ## Next Steps | 後續步驟

package/bundled/skills/bdd-assistant/SKILL.md

@@ -4,6 +4,7 @@ scope: partial
 description: "[UDS] Guide through Behavior-Driven Development workflow"
 allowed-tools: Read, Write, Grep, Glob
 argument-hint: "[feature or spec | 功能或規格]"
+status: reference
 ---
 <!-- DEPRECATION NOTICE (XSPEC-086 Phase 4, 2026-04-28):
   BDD lifecycle orchestration (4-phase cycle, Gherkin scaffolding, RED/GREEN execution) relocated to
@@ -14,6 +15,9 @@ argument-hint: "[feature or spec | 功能或規格]"
 
 # BDD Assistant | BDD 助手
 
+> ⚠️ **Status: Reference（參考用，非可執行流程）** — BDD 生命週期協調（4 階段循環、Gherkin scaffolding、RED/GREEN 執行）已於 2026-04-28 移至 **adoption layer（XSPEC-095）**。此處保留 Gherkin 格式定義、Three Amigos 結構、BDD 循環圖供參考。
+> Lifecycle orchestration has moved to the adoption layer (XSPEC-095); this document is **reference-only**. For enforced execution, use your adoption layer's toolchain.
+
 Guide through the Behavior-Driven Development (BDD) workflow using Given-When-Then format.
 
 引導行為驅動開發（BDD）流程，使用 Given-When-Then 格式。
@@ -28,6 +32,9 @@ DISCOVERY ──► FORMULATION ──► AUTOMATION ──► LIVING DOCS
 
 ## Workflow | 工作流程
 
+> 📖 以下為流程的**參考結構**，非可執行步驟；自動化／強制執行請用 adoption layer 工具鏈（XSPEC-095）。
+> The steps below are a **reference structure**, not an executable workflow.
+
 ### 1. DISCOVERY - Explore Behavior | 探索行為
 Discuss with stakeholders, identify examples and edge cases, understand the "why".
 

package/bundled/skills/bdd-assistant/guide.md

@@ -193,11 +193,10 @@ When working with Spec-Driven Development:
 ```gherkin
 # Reference spec in feature file
 # @spec SPEC-001
-@SPEC-001
 Feature: User Authentication
   Implements SPEC-001 user authentication requirements.
 
-  @AC-1
+  @SPEC-001 @AC-1
   Scenario: Successful login
     # Acceptance Criterion 1 from SPEC-001
     ...

package/bundled/skills/brainstorm-assistant/SKILL.md

@@ -14,13 +14,17 @@ Structured ideation before specification writing. Transform vague ideas into act
 
 在撰寫規格前進行結構化發想。以 2024–2026 年 AI 輔助發想研究為基礎，透過引導式腦力激盪，將模糊構想轉化為可執行的功能提案。
 
-> **Implements**: XSPEC-247 brainstorm v3 — Multi-Persona Ensemble + Multi-Critic Convergence
-> (supersedes the v2 "cognitive science upgrade").
+> **Implements**: XSPEC-296 Brainstorm Quality Standard (BQS v1) — brainstorm v4, layered on
+> XSPEC-247 brainstorm v3 (Multi-Persona Ensemble + Multi-Critic Convergence).
 
 **What changed in v3 | v3 的核心改動:** Divergence is no longer a single AI voice racing to a count — it is a **persona ensemble** (each role reasons via chain-of-thought, in isolation) crossed with **diversity lenses**. Convergence is no longer one AI scorer plus one devil's-advocate — it is a **multi-critic panel** plus a **hard-role rebuttal** (Devil's Advocate + Steelman). This directly targets the strongest finding in the literature: multiple personas beat a single pass, and a single LLM critic is weak and prone to sycophancy.
 
 v3 把發散從「單一 AI 衝數量」改為**persona 集成**（每個角色以思維鏈獨立推理）× **多樣性透鏡**；把收斂從「單一 AI 評分 + 單一反駁」改為**多評審面板** + **硬角色反駁**（Devil's Advocate + Steelman）。直接對應文獻最強結論：多 persona 勝過單一 pass，單一 LLM 評審既弱又易諂媚。
 
+**What changed in v4 | v4 的核心改動:** v3 supplied strong *mechanisms* but no decidable pass/fail quality gate. v4 layers a **time-sequenced quality contract** on top — the **Brainstorm Quality Standard (BQS v1)** — without removing any v3 behaviour. BQS is a **four-layer × timeline** structure: Layer 0 declares explore/exploit intent; Layer 1 (process, leading, visible during divergence) runs dimensions **D1–D4**; Layer 2 (product, leading, applied to Top 3 only after convergence) runs **D5–D8** plus a Seeds column and a contested zone; Layer 3 is an ungoverned **Judgment Override**. First principle: **decisions use leading signals; calibration uses lagging signals — these are ordered in time, not a right/wrong trade-off.** The hard sequence gate forbids D5–D8 from being revealed or scored during divergence. See "BQS v1 — Quality Contract" below.
+
+**v4 的核心改動：** v3 提供了強健的*機制*，卻沒有可判定 pass/fail 的品質閘。v4 在其上疊加一道**時序化品質契約**——**腦力激盪品質標準（BQS v1）**——且不移除任何 v3 行為。BQS 是**四層 × 時間軸**結構：第 0 層宣告 explore/exploit 意圖；第 1 層（過程、leading、發散期全程可見）跑維度 **D1–D4**；第 2 層（產物、leading、收斂後僅施於 Top 3）跑 **D5–D8** 加 Seeds 欄與爭議區；第 3 層是不受治理的 **Judgment Override**。第一原理：**決策用 leading 訊號、校準用 lagging 訊號——兩者是時間前後，而非對錯取捨。** 硬序列閘禁止 D5–D8 在發散期被揭示或評分。詳見下方「BQS v1 — 品質契約」。
+
 ## Mode Selection | 使用前先選模式
 
 Apply these **objective triggers** before starting. Default is full v3 — routing rules are shortcuts to skip phases, not barriers to add.
@@ -45,10 +49,79 @@ Apply these **objective triggers** before starting. Default is full v3 — routi
 ```
 [Mode Selection] ─► PRE-FLIGHT ─► FRAME ─► DIVERGE ───────────► CONVERGE ──────────► OUTPUT
    客觀路由          防止錨定      定義問題   persona 集成+透鏡       多評審面板+硬角色反駁    輸出提案
+   ▲ Layer 0 intent              ▲ Layer 1 (D1–D4, leading)        ▲ Layer 2 (D5–D8, Top 3)   ▲ Layer 3 override
 ```
 
 ---
 
+## BQS v1 — Quality Contract | BQS v1 — 品質契約
+
+> First principle | 第一原理: A brainstorm's output is a **hypothesis, not an answer**. Whether an idea is good is unknowable at ideation time, so quality can only be judged on **leading** signals (process + epistemic integrity); the standard's legitimacy is **calibrated** by **lagging** signals (later outcomes). Decisions use leading; calibration uses lagging; they are ordered in **time**, not a right/wrong choice. (This corrects any "use only leading, never lagging" absolutism.)
+>
+> 腦力激盪的產物是**假說，不是答案**。一個點子好不好在發想當下不可知，所以品質只能用 **leading** 訊號（過程＋認識論完整性）判；標準的正當性靠 **lagging** 訊號（事後結果）**校準**。決策用 leading、校準用 lagging，兩者是**時間**前後，不是對錯取捨。（此處修正任何「只用 leading 不用 lagging」的絕對說法。）
+
+BQS is a **four-layer × timeline** contract. It is **additive** to v3 — every v3 flag and mechanism is preserved (see "Backward Compatibility").
+
+BQS 是**四層 × 時間軸**契約。它是 v3 的**疊加**——所有 v3 旗標與機制都保留（見「向後相容」）。
+
+### Layer 0 — Intent (opening declaration) | 第 0 層 — 意圖（開場宣告）
+
+At the start, declare the **explore/exploit ratio** and the bet type (incremental vs barbell long-tail). This layer **modulates dimension weights**: under an exploit-leaning session, a **low D2 (divergence coverage) score is correct and is NOT penalised**.
+
+開場宣告本次 **explore/exploit 配比** 與賭注類型（漸進 vs barbell 長尾）。此層**調節維度權重**：偏 exploit 的工作階段，**D2（發散覆蓋）低分是正確的、不扣分**。
+
+### Layer 1 — Process (divergence-period leading, fully visible) | 第 1 層 — 過程（發散期 leading，全程可見）
+
+| Dim | Oracle (how it fails) | Oracle（怎麼判 fail）|
+|-----|-----------------------|----------------------|
+| **D1 Frame purity** | Problem embeds a specific solution or a "like X but for Y" framing → fail; must run 5-Whys to the root cause | 問題內嵌特定方案或「像 X 給 Y」→fail；須 5-Whys 到根因 |
+| **D2 Divergence coverage** | Surviving ideas span < 3 independent personas/lenses → fail (modulated by Layer 0 weight); only a round with zero new ideas counts as saturation | 存活想法跨 <3 獨立 persona/透鏡→fail（受第 0 層權重調節）；連一輪零新增才算飽和 |
+| **D3 Cross-session diversity** | Seed is a competitor analogy → fail; Top all from a single lens → fail | 種子是競品類比→fail；Top 全來自單一 lens→fail |
+| **D4 Evaluation de-bias** | Requires ≥3 critics + hard-role Devil's Advocate + Steelman; **pass requires an independent context**, otherwise mark `[degraded]` — must NOT be marked pass | 須 ≥3 評審 + 硬角色 Devil's Advocate + Steelman；**pass 須在獨立 context 執行**，否則標 `[degraded]` 不得標 pass |
+
+> **Hard sequence gate | 硬序列閘:** D5–D8 are **forbidden** from being revealed or scored during the divergence period. The CONVERGE critics MUST NOT be invoked before the **last persona** has produced its set.
+>
+> D5–D8 禁止在發散期揭示或評分；CONVERGE 的 critic 不得在**最後一個 persona** 產完前被呼叫。
+
+### Layer 2 — Product (post-convergence leading, applied to Top 3 only) | 第 2 層 — 產物（收斂後 leading，僅施於 Top 3）
+
+> Evaluative dimensions are confined to **after convergence × Top 3 only** — never as a full gate on every divergence idea (that would retroactively pollute divergence, kill evidence-free future ideas, and create form-filling theatre). | 評判維度限縮在**收斂後 × 僅 Top 3**，絕不在發散期對全部想法當硬閘（否則回溯污染發散、扼殺無證據的未來想法、製造填表劇場）。
+
+| Dim | Oracle | Oracle |
+|-----|--------|--------|
+| **D5 Grounding** | A `[current state / external fact]` claim with no file:line/source → fail; a `[future / hypothesis]` claim needs no grounding, marked `[hypothesis]`. **An external-fact claim is a cross-tier floor (applies even at the creative tier).** | 「現狀/外部事實」主張無 file:line/來源→fail；「未來/假說」免接地、標 `[假說]`。**外部事實宣稱為跨級地板**（creative 級亦適用）|
+| **D6 Net benefit** | A selected idea that does not answer "whose problem / do we actually have it / cost of not doing it" → fail; at least one "not worth doing" elimination required. **Attach a lagging registry field**: which signal will later validate this judgement? | 入選想法未答「解誰問題／我們真有嗎／不做的代價」→fail；至少一個「不值得做」淘汰。**掛 lagging 登記欄**：此判斷事後拿什麼訊號驗證？ |
+| **D7 Falsifiability** | Two-state: `[falsifiable now]` or `[need to do X first to define falsification]`; the latter is **routed to a next-step feeding D8 and does NOT count as fail** | 二態：`[現可陳述證偽]` 或 `[需先做 X 才能定義證偽]`；後者**轉 next-step 餵 D8，不算 fail** |
+| **D8 Actionability** | No next-step decision (including "defer") → fail | 無 next-step 裁決（含「暫不做」）→fail |
+| **+ Seeds column** | When killing an idea, force-store "what was wrong, what real problem it points to"; if ≥1 idea was killed then Seeds ≥1 (only checked for non-empty) | 殺想法時強制存「錯在哪、指向什麼真問題」；被殺 ≥1 則 Seeds ≥1（只檢查非空）|
+| **+ Contested zone** | Ideas whose critic variance exceeds a threshold are surfaced in a split view and **not eliminated by mean ranking** (protects barbell long-tail) | critic 方差 > 閾值的想法分流呈現，不按 mean 排序淘汰（守 barbell 長尾）|
+
+### Layer 3 — Ungoverned zone (Judgment Override) | 第 3 層 — 不可治理區（Judgment Override）
+
+A reserved space no oracle enters: a human's intuition to "keep / kill" needs only a one-line reason and **overrides the aggregate score — it passes through no dimension**. This admits the standard is incomplete and avoids "brainstorming to pass dimensions".
+
+明文留一塊 oracle 不進入的地：人類直覺「保留/斃掉」附一句理由即可，**凌駕聚合分數、不需過任何維度**。承認標準不完備，避免「為過維度而 brainstorm」。
+
+### Structural rules | 結構規則
+
+1. **Meta stop rule | Meta 停止規則:** stop when the layer's dimensions are all green **and**, after one more round, the **Top 3 set membership is unchanged** (set membership, not internal ranking). **Hard cap: 2 rounds.** This replaces the vague "decision didn't flip" criterion. | 該層維度全綠 **且** 再跑一輪後 **Top 3 集合成員不變**（看集合成員、不看內部排序）→停。**硬上限 2 輪**。取代模糊的「不翻決策」。
+2. **Judge ≠ generator | 判官≠產生者:** D2/D4/D5/D7 judgements need an independent viewpoint; single-context self-evaluation may only be `[degraded]`, never pass. | D2/D4/D5/D7 判定須獨立視角；單 context 自評只能 `[degraded]`，不得 pass。
+3. **Calibration loop (consumes lagging) | 校準回路（吃 lagging）:** v3's three Session Self-Evaluation metrics are folded in as the **lagging** end — **Adoption Rate = D6 lagging validation, Diversity = D2/D3 lagging observation, Cognitive Load = a cost constraint**. **Two parallel evaluation systems are forbidden** — there is one loop, not a separate self-eval. | v3 的 **Adoption Rate**＝D6 滯後驗證、**Diversity**＝D2/D3 滯後觀測、**Cognitive Load**＝成本約束；三者收編為此回路 lagging 端，**禁兩套平行評估**。
+4. **BQS self-evolution (lightweight) | BQS 自我演化（輕量）:** versioned + an evidence `last-reviewed` that flags when overdue; "periodically brainstorm BQS itself" is optional. | 版本化 + 證據清單 last-reviewed 逾期 flag；「定期對 BQS 自身 brainstorm」列可選。
+5. **Minimal sufficiency | 最小充分原則:** apply only the fewest dimensions the tier requires + the Top-3 confinement — this is the gate for cognitive economy; no separate cost dimension is added. | 套用該 tier 所需的最少維度 + Top-3 限縮——認知經濟性的守門，不另設成本維度。
+
+### Tiering (bound to v3 objective triggers, not self-assessment) | 分級（綁 v3 既有客觀觸發，非自評）
+
+| Tier (from Mode Selection triggers) | BQS dimensions applied | 套用維度 |
+|-------------------------------------|------------------------|----------|
+| creative / `--quick` | D1–D3 **+ D5 external-fact floor** | D1–D3 ＋ D5 外部事實地板 |
+| default | D1–D5 + D8 | D1–D5 + D8 |
+| strategic / architecture / business | all layers | 全層 |
+
+> Tiers are selected by the **objective Mode Selection triggers** (word count, flags, existence of a spec) — not by the initiator's self-assessment. | 分級由**客觀模式選擇觸發表**（字數、旗標、是否有規格）決定，而非發起者自評。
+
+---
+
 ### Phase 0: PRE-FLIGHT | 防止 AI 錨定
 
 **Why this phase exists:** Independent ideas written *before* the AI generates anything consistently produce more diverse results. In AI-assisted contexts this matters *more*, not less: research on design fixation shows that AI output — being fluent and high-fidelity — **deepens** fixation rather than relieving it (Wadinambiarachchi et al., CHI 2024).
@@ -73,6 +146,10 @@ Apply these **objective triggers** before starting. Default is full v3 — routi
 >
 > **反種子 guardrail（v3 新增）：** 不要用「像 X 但給 Y」的框架當種子（如「給醫生用的 Slack」）。這類產品類比種子會把 LLM 鎖進單一解空間、明顯降低想法多樣性。請捕捉底層**問題**，而非產品類比。
 
+> **BQS Layer 0 — Intent declaration (new in v4):** Before FRAME, state the **explore/exploit ratio** and bet type (incremental vs barbell long-tail) for this session. This modulates D2 weighting downstream — under an exploit-leaning session, low divergence coverage is correct and is not penalised. Default when unstated: explore-leaning.
+>
+> **BQS 第 0 層 — 意圖宣告（v4 新增）：** 在 FRAME 之前，宣告本次的 **explore/exploit 配比** 與賭注類型（漸進 vs barbell 長尾）。此宣告下游調節 D2 權重——偏 exploit 時低發散覆蓋是正確的、不扣分。未宣告時預設：偏 explore。
+
 **Flag:** `--skip-preflight` bypasses this phase with a one-line warning:
 `⚠ Skipping Pre-flight may cause AI anchoring`
 
@@ -137,6 +214,10 @@ The "best ideas appear in the second half" pattern (Nijstad) is a **human-group*
 
 「好點子在後半」（Nijstad）是**人類群體**現象，**未在 LLM 證實**（LLM 多為高原/枯竭）。故固定數量門檻降為**輔助提示**：若全組少於約 8 個相異想法，提示「繼續——加一個還沒用過的 persona 或透鏡」。真正的門檻是**多樣性**（覆蓋了幾個不同視角），而非數量。
 
+> **BQS Layer 1 applies here (D1–D4, leading, visible) | 此處生效 BQS 第 1 層（D1–D4，leading，全程可見）:** during divergence, only **D1 (frame purity), D2 (divergence coverage, weighted by Layer 0), D3 (cross-session diversity), D4 (evaluation de-bias)** are scored and shown. **Hard sequence gate:** the evaluative dimensions **D5–D8 must NOT be revealed or scored before the last persona has produced its set** — the CONVERGE critics are not invoked until divergence is complete.
+>
+> 發散期只對 **D1（框架純度）、D2（發散覆蓋，依第 0 層加權）、D3（跨會話多樣性）、D4（評估去偏）** 評分與呈現。**硬序列閘：** 評判維度 **D5–D8 在最後一個 persona 產完前不得揭示或評分**——CONVERGE 的 critic 在發散完成前不被呼叫。
+
 #### Classic techniques (still available) | 經典技法（仍保留）
 
 | Technique | When to Use | 使用時機 |
@@ -193,6 +274,25 @@ The user **must** respond to each before advancing:
 
 **Flag:** `--no-rebuttal` skips this step; report section marked "Rebuttal: skipped".
 
+> **BQS D4 — judge ≠ generator | BQS D4 — 判官≠產生者:** D4 passes **only when the critics/Devil's Advocate run in an independent context** (the `--enhanced` isolated-agent host). On a baseline single context, the panel runs as same-context self-evaluation, which is marked `[degraded]` and **must NOT be marked pass** — it is honest but not independent. Do not silently treat a baseline run as a D4 pass.
+>
+> D4 唯有**評審/Devil's Advocate 在獨立 context 執行**（`--enhanced` 隔離 agent 宿主）才可 pass。baseline 單 context 下，面板是同 context 自評，標 `[degraded]`、**不得標 pass**——誠實但非獨立。不得把 baseline 跑當成 D4 pass 靜默通過。
+
+#### Step 3c: BQS Layer 2 — product gate on Top 3 | BQS 第 2 層——對 Top 3 的產物閘
+
+After convergence, apply **D5–D8 to the Top 3 only** (never to every divergence idea). For each Top-3 idea:
+
+收斂後，**僅對 Top 3** 套用 **D5–D8**（絕不對全部發散想法）。對每個 Top-3 想法：
+
+- **D5 Grounding:** split each claim into `[current state / external fact]` (needs a file:line or source, else fail) vs `[future / hypothesis]` (no grounding needed, mark `[hypothesis]`). An **external-fact claim is a cross-tier floor** — it must be grounded even at the creative tier. | 將每個主張分流為 `[現狀/外部事實]`（需 file:line 或來源，否則 fail）vs `[未來/假說]`（免接地、標 `[假說]`）。**外部事實宣稱為跨級地板**——creative 級也須接地。
+- **D6 Net benefit:** each Top-3 idea must answer "whose problem / do we actually have it / cost of not doing it"; at least one idea must be eliminated as "not worth doing"; attach a **lagging registry field** (which later signal validates this). | 每個 Top-3 須答「解誰問題／我們真有嗎／不做的代價」；至少一個淘汰為「不值得做」；掛 **lagging 登記欄**（事後哪個訊號驗證）。
+- **D7 Falsifiability (two-state):** mark `[falsifiable now]` **or** `[need to do X first to define falsification]`. The latter is **routed to a next-step that feeds D8 — it does NOT count as fail**. | 標 `[現可陳述證偽]` **或** `[需先做 X 才能定義證偽]`。後者**轉 next-step 餵 D8——不算 fail**。
+- **D8 Actionability:** every surviving idea needs a next-step decision (including an explicit "defer / not now"); none → fail. | 每個存活想法需 next-step 裁決（含明確「暫不做」）；無 → fail。
+
+> **Meta stop rule (BQS structural rule 1) | Meta 停止規則:** stop iterating when the applied dimensions are all green **and** one more round leaves the **Top 3 set membership unchanged** (set membership, not internal ordering). **Hard cap: 2 rounds.**
+>
+> 當套用的維度全綠 **且** 再跑一輪後 **Top 3 集合成員不變**（看集合成員、非內部排序）→停。**硬上限 2 輪。**
+
 ---
 
 ### Phase 4: OUTPUT | 輸出提案
@@ -217,8 +317,21 @@ Produce a Brainstorm Report ready for `/requirement` or `/sdd`. Each surviving i
 |---|------|---------|------|-------------|---------------|--------------|-----------|
 | 1 | ...  | Skeptic | Reversal | 4.0 | 4.5 | 4.0 | 4.2 |
 
-## Top 3 Recommendations
+## Top 3 Recommendations (BQS Layer 2 applied)
 1. **[Idea]** ✓ Passed rebuttal — [Why] — Persona: [..] — [User rebuttal response]
+   - D5 grounding: [current-state claims with file:line | future claims marked [hypothesis]]
+   - D6 net benefit: [whose problem / do we have it / cost of not doing] — lagging signal: [..]
+   - D7 falsifiability: [falsifiable now | need to do X first → next-step]
+   - D8 next-step: [action | defer]
+
+## Contested Zone (high critic-variance ideas)
+[Ideas whose critic variance exceeded threshold — surfaced, NOT eliminated by mean ranking (barbell long-tail)]
+
+## Seeds (from killed ideas)
+[For each killed idea: what was wrong, what real problem it points to. Non-empty if ≥1 idea was killed.]
+
+## Judgment Override (Layer 3, ungoverned)
+[Human keep/kill decisions that override the aggregate score, each with a one-line reason. Optional.]
 
 ## Diversity Note
 [How many distinct lenses/personas the surviving ideas span — flag if all from one cluster]
@@ -231,6 +344,10 @@ Produce a Brainstorm Report ready for `/requirement` or `/sdd`. Each surviving i
 - [ ] Proceed to `/sdd` if requirements are clear
 ```
 
+> **BQS output additions (v4):** the **Seeds** section (rule: non-empty when ≥1 idea was killed), the **Contested Zone** (high-variance ideas not eliminated by mean ranking), and the **Judgment Override** channel (human decision overriding the aggregate score) are required by BQS Layer 2/3. The Top-3 block records the D5–D8 status per idea.
+>
+> **BQS 輸出新增（v4）：** **Seeds** 區（規則：被殺 ≥1 則非空）、**爭議區**（高方差想法不按 mean 淘汰）、**Judgment Override** 通道（人類裁決凌駕聚合分）為 BQS 第 2/3 層所要求。Top-3 區塊逐項記錄 D5–D8 狀態。
+
 ## Diversity-Collapse Guardrail | 多樣性崩塌防護
 
 Using a single LLM for ideation reduces the **diversity of ideas across users**, even when each individual feels more creative (Anderson, Shah & Kreminski, 2024; corroborated by the widely-cited Doshi & Hauser, *Science Advances* 2024). Guard against it:
@@ -263,17 +380,21 @@ Multi-agent ideation (independent agents conversing/contributing) outperforms a
 | **Devil's Advocate + Steelman** | Hard-role rebuttal | 硬角色反駁 |
 | **SCAMPER / Six Hats** | Classic divergence (usable as personas) | 經典發散（可當 persona） |
 
-## Session Self-Evaluation | 工作階段自評
+## Calibration Loop — Lagging Signals (BQS) | 校準回路——lagging 訊號（BQS）
+
+> **This is the lagging end of the single BQS quality loop — NOT a second, parallel evaluation system.** The leading decision is made by BQS Layers 0–2 during the session; these three metrics calibrate the standard afterwards. Do not run a separate "self-evaluation" alongside BQS.
+>
+> **這是同一道 BQS 品質回路的 lagging 端——不是第二套平行評估系統。** 工作階段中由 BQS 第 0–2 層做 leading 決策；這三個指標事後校準標準。禁止在 BQS 之外另跑一套「自評」。
 
-After each session, record three metrics (1–5) to track improvement over time.
+After each session, record three metrics (1–5). Each maps onto a BQS dimension as its lagging validation:
 
-每次工作階段結束後記錄三個指標（1–5 分），追蹤長期改善。
+每次工作階段結束後記錄三個指標（1–5 分），各自對應一個 BQS 維度作為其滯後驗證：
 
-| Metric | Question | 指標 |
-|--------|----------|------|
-| **Adoption Rate** | How many of today's ideas will I actually use? | 採用率 |
-| **Diversity** | Did surviving ideas span multiple personas/lenses? | 多樣性（跨幾個 persona/透鏡） |
-| **Cognitive Load** | How mentally taxing was this? (5 = effortless) | 認知負擔 |
+| Metric | Question | BQS lagging role | 指標 / 滯後角色 |
+|--------|----------|------------------|------------------|
+| **Adoption Rate** | How many of today's ideas will I actually use? | **D6 net-benefit lagging validation** | 採用率 = D6 淨值滯後驗證 |
+| **Diversity** | Did surviving ideas span multiple personas/lenses? | **D2/D3 lagging observation** | 多樣性 = D2/D3 滯後觀測 |
+| **Cognitive Load** | How mentally taxing was this? (5 = effortless) | **cost constraint** | 認知負擔 = 成本約束 |
 
 Collect 3 sessions before drawing conclusions. See [guide.md](./guide.md) for the full A/B experiment protocol.
 
@@ -283,6 +404,7 @@ Collect 3 sessions before drawing conclusions. See [guide.md](./guide.md) for th
 
 | Flag | Effect | 說明 |
 |------|--------|------|
+| `--intent explore\|exploit\|<ratio>` | Declare BQS Layer 0 explore/exploit intent (modulates D2 weight) | 宣告 BQS 第 0 層 explore/exploit 意圖（調節 D2 權重） |
 | `--personas "a,b,c"` | Override the default persona ensemble | 覆寫預設 persona 組 |
 | `--lens analogical\|reversal\|morphological` | Force a primary diversity lens | 指定主要多樣性透鏡 |
 | `--enhanced` | Parallel persona/critic agents (falls back if unsupported) | 平行 persona/評審 agent（不支援則退回） |

package/bundled/skills/brainstorm-assistant/guide.md

@@ -10,8 +10,8 @@ description: |
 
 > **Language**: English | [繁體中文](../../locales/zh-TW/skills/brainstorm-assistant/guide.md)
 
-**Version**: 3.0.0
-**Last Updated**: 2026-06-01
+**Version**: 4.0.0
+**Last Updated**: 2026-06-22
 **Applicability**: All software projects
 **Scope**: universal
 **Type**: Utility Skill (no core standard)
@@ -43,6 +43,17 @@ v1 was a generic FRAME→DIVERGE→CONVERGE flow. v2 added cognitive-science gat
 
 The pre-flight phase is **kept and strengthened** (fixation research says AI anchoring is real and possibly worse), while the 10-idea gate and the single-AI rebuttal are **demoted / hardened** because their original human-group evidence does not transfer cleanly to a single LLM.
 
+### What v4 changes (BQS v1)
+
+v3 supplied strong *mechanisms* but no **decidable pass/fail quality gate** — its only close analogue, the Session Self-Evaluation, was after-the-fact, self-assessed, and non-blocking. **v4 layers the Brainstorm Quality Standard (BQS v1) on top of v3 without removing anything** (see "Backward Compatibility"). BQS is a **four-layer × timeline** quality contract:
+
+- **Layer 0 — Intent**: declare explore/exploit ratio; modulates dimension weights.
+- **Layer 1 — Process (divergence-period leading, visible)**: dimensions **D1 frame purity / D2 divergence coverage / D3 cross-session diversity / D4 evaluation de-bias**, each with an oracle. A **hard sequence gate** forbids the evaluative dimensions D5–D8 from being revealed or scored during divergence.
+- **Layer 2 — Product (post-convergence leading, Top 3 only)**: **D5 grounding / D6 net benefit / D7 falsifiability / D8 actionability**, plus a Seeds column and a contested zone for high-variance ideas.
+- **Layer 3 — Ungoverned**: a Judgment Override where human intuition overrides the aggregate score.
+
+The first principle is refined: **decisions use leading signals, calibration uses lagging signals — ordered in time, not a right/wrong trade-off** (this removes any "use only leading, never lagging" absolutism). The core claim: evaluative dimensions are confined to **after convergence × Top 3 only**, never as a full gate on every divergence idea — otherwise they retroactively pollute divergence, kill evidence-free future ideas, and create form-filling theatre. The full contract, oracles, and structural rules live in **SKILL.md → "BQS v1 — Quality Contract"**.
+
 ---
 
 ## Research Foundations | 認知科學依據
@@ -399,11 +410,15 @@ The Mode Selection table in SKILL.md uses objective triggers (word count, presen
 
 ---
 
-## Self-Evaluation Framework | 自我評估框架
+## Calibration Loop (BQS lagging end) | 校準回路（BQS lagging 端）
+
+> **In v4 these three metrics are the lagging end of the single BQS quality loop, not a second parallel evaluation system.** The leading pass/fail decision is made by BQS Layers 0–2 during the session; these metrics calibrate the standard afterwards. **Adoption Rate = D6 net-benefit lagging validation, Diversity = D2/D3 lagging observation, Cognitive Load = a cost constraint.** Do not run a separate self-evaluation alongside BQS.
+>
+> **v4 中這三個指標是同一道 BQS 品質回路的 lagging 端，不是第二套平行評估系統。** 工作階段中由 BQS 第 0–2 層做 leading pass/fail 決策；這些指標事後校準標準。**採用率＝D6 淨值滯後驗證、多樣性＝D2/D3 滯後觀測、認知負擔＝成本約束。** 禁止在 BQS 之外另跑一套自評。
 
-Record three metrics after every session to build an empirical record. Do not judge v3 vs v2 on a single session — collect at least 3 comparable sessions.
+Record three metrics after every session to build an empirical record. Do not judge across versions on a single session — collect at least 3 comparable sessions.
 
-每次工作階段後記錄三個指標，建立實證紀錄。不要以單次評估 v3 vs v2，至少收集 3 次可比較的工作階段。
+每次工作階段後記錄三個指標，建立實證紀錄。不要以單次評估跨版本，至少收集 3 次可比較的工作階段。
 
 ### The Three Metrics
 
@@ -538,6 +553,18 @@ After 6 weeks, review session logs and A/B data to calibrate which combination f
 
 ---
 
+## Backward Compatibility (v3 → v4) | 向後相容
+
+BQS v1 is an **additive quality contract**, not a rewrite. Everything from v3 is preserved:
+
+- **All flags**: `--personas`, `--lens`, `--enhanced`, `--skip-preflight`, `--no-rebuttal`, `--quick`, `--technique` behave exactly as in v3. v4 adds one optional flag, `--intent`, for the Layer 0 declaration.
+- **All mechanisms**: PRE-FLIGHT anti-anchoring, FRAME 5-Whys/HMW, the persona ensemble, diversity lenses, the multi-critic panel, the hard-role rebuttal (Devil's Advocate + Steelman), the Diversity-Collapse Guardrail, and the Enhanced Tier are unchanged.
+- **The `--enhanced` isolated-agent host** is what lets BQS **D4 pass** (judge ≠ generator); on a baseline single context the panel is marked `[degraded]`, never silently passed.
+
+BQS v1 是**疊加的品質契約**，不是打掉重練。v3 的一切都保留：所有旗標（`--personas`/`--lens`/`--enhanced`/`--skip-preflight`/`--no-rebuttal`/`--quick`/`--technique`）行為不變，v4 僅新增一個可選的 `--intent` 旗標供第 0 層宣告；所有機制（PRE-FLIGHT 防錨定、FRAME 5-Whys/HMW、persona 集成、多樣性透鏡、多評審面板、硬角色反駁、多樣性崩塌防護、Enhanced 層）不變；`--enhanced` 的隔離 agent 宿主正是讓 BQS **D4 pass**（判官≠產生者）的條件，baseline 單 context 標 `[degraded]`、不靜默通過。
+
+---
+
 ## Integration with UDS Workflow
 
 ### Mapping to `/requirement`
@@ -576,6 +603,7 @@ After 6 weeks, review session logs and A/B data to calibrate which combination f
 
 | Version | Date | Changes |
 |---------|------|---------|
+| 4.0.0 | 2026-06-22 | XSPEC-296: Brainstorm Quality Standard (BQS v1) — a four-layer × timeline quality contract layered additively on v3. Layer 0 explore/exploit intent (modulates D2 weight); Layer 1 process leading dimensions D1–D4 with a hard sequence gate (D5–D8 forbidden during divergence); Layer 2 product leading dimensions D5–D8 applied to Top 3 only, with a Seeds column and high-variance contested zone; Layer 3 Judgment Override (overrides aggregate score). D4 judge≠generator (independent context else `[degraded]`); D5 claim split (external-fact floor cross-tier); D7 two-state falsifiability ("need to do X" → next-step, not fail); Meta stop rule (Top-3 set stability + hard cap 2 rounds). Session Self-Evaluation re-positioned as the calibration loop's lagging end (Adoption→D6, Diversity→D2/D3, Cognitive Load→cost; two parallel evaluations forbidden). First principle refined to "decisions use leading, calibration uses lagging". Tiering bound to objective Mode Selection triggers. New flag `--intent`. All v3 flags/mechanisms preserved. |
 | 3.0.0 | 2026-06-01 | XSPEC-247: DIVERGE re-centred on persona ensemble + diversity lenses (analogical/reversal/morphological); CONVERGE re-centred on multi-critic panel + hard-role rebuttal (Devil's Advocate + Steelman); Diversity-Collapse Guardrail; Enhanced Tier (parallel persona/critic agents, graceful fallback); Research Foundations rebuilt on 6 verified 2024–2026 sources; Validity Caveats re-rated (pre-flight LOW, Nijstad/Nemeth demoted); new flags `--personas`/`--lens`/`--enhanced`; anti-seed guardrail |
 | 2.1.0 | 2026-05-09 | XSPEC-196 Phase 2: Mode Selection objective routing; Self-Evaluation Framework; A/B Experiment Protocol; Research Validity Caveats; Gradual Adoption Protocol |
 | 2.0.0 | 2026-05-09 | XSPEC-196: Phase 0 Pre-flight (anti-anchoring), Rebuttal Round, 10-idea minimum gate + semantic batching |

package/bundled/skills/checkin-assistant/SKILL.md

@@ -4,6 +4,7 @@ scope: partial
 description: "[UDS] Pre-commit quality gates verification"
 allowed-tools: Read, Grep, Glob, Bash(git diff:*), Bash(git status:*), Bash(npm test:*), Bash(npm run lint:*)
 disable-model-invocation: true
+status: reference
 ---
 <!-- DEPRECATION NOTICE (XSPEC-086 Phase 4, 2026-04-28):
   Checkin workflow orchestration (gate execution sequence, abort logic) relocated to adoption layer
@@ -14,12 +15,18 @@ disable-model-invocation: true
 
 # Check-in Assistant | 簽入助手
 
+> ⚠️ **Status: Reference（參考用，非可執行流程）** — 簽入流程協調（關卡執行序列、abort 邏輯）已於 2026-04-28 移至 **adoption layer（XSPEC-095）**。此處保留品質關卡定義、檢查清單項、never-commit 規則供參考。
+> Lifecycle orchestration has moved to the adoption layer (XSPEC-095); this document is **reference-only**. For enforced execution, use your adoption layer's toolchain.
+
 Verify pre-commit quality gates before committing code to ensure codebase stability.
 
 在提交程式碼前驗證品質關卡，確保程式碼庫的穩定性。
 
 ## Workflow | 工作流程
 
+> 📖 以下為流程的**參考結構**，非可執行步驟；自動化／強制執行請用 adoption layer 工具鏈（XSPEC-095）。
+> The steps below are a **reference structure**, not an executable workflow.
+
 1. **Check git status** - Run `git status` and `git diff` to understand pending changes
 2. **Run tests** - Execute `npm test` (or project test command) to verify all tests pass
 3. **Run linting** - Execute `npm run lint` to check code style compliance
@@ -60,7 +67,7 @@ After `/checkin` completes, the AI assistant should suggest:
 > **品質關卡驗證完成。建議下一步 / Quality gate verification complete. Suggested next steps:**
 > - 全部通過 ✅ → 執行 `/commit` 提交變更 ⭐ **Recommended / 推薦** — All passed → Run `/commit` to commit
 > - 有失敗項目 ❌ → 修復問題後重新執行 `/checkin` — Failures found → Fix then re-run `/checkin`
-> - 需要程式碼審查 → 執行 `/review` 進行自我審查 — Need review → Run `/review` for self-review
+> - 需要程式碼審查 → 執行 `/code-review` 進行自我審查 — Need review → Run `/code-review` for self-review
 > - UDS 安裝有異常 → 執行 `/audit` 診斷問題 — UDS issues detected → Run `/audit` to diagnose
 
 ## Reference | 參考

package/bundled/skills/ci-cd-assistant/SKILL.md

@@ -62,6 +62,62 @@ BUILD ──► TEST ──► ANALYZE ──► DEPLOY ──► VERIFY
 | **GitLab CI** | `cache:` keyword | `parallel:` keyword | CI/CD Variables | 使用 `include:` 模組化 |
 | **Jenkins** | Stash/unstash | `parallel {}` block | Credentials plugin | 使用 shared libraries |
 
+## CI Job Orchestration Patterns | CI 作業編排模式
+
+Recurring CI job-orchestration patterns (UDS #126). Each principle is platform-agnostic; the mapping columns show how each platform expresses it.
+反覆出現的 CI 作業編排模式（UDS #126）。每條原則平台無關；對照欄顯示各平台的寫法。
+
+### Trigger separation | 觸發分離
+
+Separate heavy, slow suites (a full E2E run) from lightweight per-change checks by **trigger**, so runner cost scales with risk, not with every push.
+將重型慢速套件（完整 E2E）以**觸發條件**與每次變更的輕量檢查分離，使 runner 成本隨風險而非每次 push 成長。
+
+| Need | GitHub Actions | GitLab CI | Jenkins |
+|------|----------------|-----------|---------|
+| Heavy suite on schedule | `on: schedule:` | `rules: if $CI_PIPELINE_SOURCE == "schedule"` | cron trigger |
+| Light checks per change | `on: pull_request` | `rules: if $CI_PIPELINE_SOURCE == "merge_request_event"` | SCM webhook |
+
+> ⚠️ Anti-pattern: full E2E on every push. 反模式：每次 push 跑完整 E2E。
+
+### Shared-resource serialization | 共享資源序列化
+
+When several jobs touch one **stateful shared resource** (a single test env / runner / deploy target), serialize them to prevent concurrent corruption.
+當多個作業觸及同一個**有狀態共享資源**（單一測試環境 / runner / 部署目標）時，序列化以防並發污染。
+
+| Need | GitHub Actions | GitLab CI | Jenkins |
+|------|----------------|-----------|---------|
+| Mutual exclusion | `concurrency: { group: … }` | `resource_group:` | `lock(resource: …)` |
+
+> ⚠️ Anti-pattern: two jobs writing the same DB/env in parallel. 反模式：兩作業並行寫同一 DB/環境。
+
+### Change-detection gating | 變更偵測閘控
+
+When `HEAD` is unchanged for a scope (or a change touches only out-of-scope paths), skip deploy/test to conserve runner time. Define "unchanged" **explicitly** (last-success SHA or path filters).
+當某範圍的 `HEAD` 自上次成功未變（或變更只觸及範圍外路徑）時，跳過 deploy/test 以節省 runner。**明確**定義「未變」（上次成功 SHA 或路徑過濾）。
+
+| Need | GitHub Actions | GitLab CI | Jenkins |
+|------|----------------|-----------|---------|
+| Skip on unchanged scope | `paths:` / `dorny/paths-filter` | `rules: changes:` / SHA compare | `changeset` condition |
+
+> ⚠️ Anti-pattern: full deploy + test for a docs-only commit. 反模式：文件-only commit 跑完整部署。
+
+### Advisory vs. gating jobs | 諮詢 vs 閘控作業
+
+Declare each job as **advisory** (failure reported, does not block merge) or **gating** (failure blocks merge). Conflating them blocks on noise or fails to gate on what matters. (Security gates are the Block/Warn/Log specialization — see [`pipeline-security-gates`](../../core/pipeline-security-gates.md).)
+明確宣告每個作業是 **advisory**（回報但不擋合併）或 **gating**（擋合併）。混淆會在雜訊上卡關、或漏掉該擋的。（安全閘門是 Block/Warn/Log 特例——見 [`pipeline-security-gates`](../../core/pipeline-security-gates.md)。）
+
+| Need | GitHub Actions | GitLab CI | Jenkins |
+|------|----------------|-----------|---------|
+| Advisory (non-blocking) | `continue-on-error: true` | `allow_failure: true` | `catchError` → `unstable` |
+| Gating (blocking) | required status check | default (no `allow_failure`) | hard fail |
+
+> ⚠️ Anti-pattern: a noisy third-party lint as a hard gate, or a critical suite as advisory. 反模式：把雜訊 lint 設硬閘，或把關鍵套件設 advisory。
+
+### Troubleshooting: `npm ci` `EUSAGE` | 排錯：`npm ci` `EUSAGE`
+
+`npm ci` requires a **committed** lockfile. If `package-lock.json` is gitignored, CI dies with `EUSAGE`. Fix: commit the lockfile, or for CI-only installs use `npm install --no-audit --no-fund --ignore-scripts` (skipping `prepare`/husky hooks not wanted in CI).
+`npm ci` 需要 **committed** lockfile。若 `package-lock.json` 被 gitignore，CI 會以 `EUSAGE` 失敗。修法：commit lockfile，或 CI-only 安裝用 `npm install --no-audit --no-fund --ignore-scripts`（跳過 CI 不需要的 `prepare`/husky hook）。
+
 ## Usage | 使用方式
 
 ```bash
@@ -90,6 +146,7 @@ After `/ci-cd` completes, the AI assistant should suggest:
 
 | Version | Date | Changes | 變更說明 |
 |---------|------|---------|----------|
+| 1.1.0 | 2026-06-24 | Add CI Job Orchestration Patterns (trigger separation, shared-resource serialization, change-detection gating, advisory-vs-gating, `npm ci` `EUSAGE` troubleshooting) — UDS #126 → XSPEC-300 | 新增 CI 作業編排模式（UDS #126） |
 | 1.0.0 | 2026-03-24 | Initial release | 初始版本 |
 
 ## License | 授權

package/bundled/skills/code-review-assistant/SKILL.md

@@ -1,10 +1,11 @@
 ---
-name: review
+name: code-review
 scope: universal
 description: "[UDS] Perform systematic code review with checklist"
 allowed-tools: Read, Grep, Glob, Bash(git diff:*), Bash(git log:*), Bash(git show:*)
 argument-hint: "[file path or branch | 檔案路徑或分支名稱]"
 disable-model-invocation: true
+status: reference
 ---
 <!-- DEPRECATION NOTICE (XSPEC-086 Phase 4, 2026-04-28):
   Code review workflow orchestration (4-step sequence, outcome gating) relocated to adoption layer
@@ -15,12 +16,18 @@ disable-model-invocation: true
 
 # Code Review Assistant | 程式碼審查助手
 
+> ⚠️ **Status: Reference（參考用，非可執行流程）** — 程式碼審查流程協調（4 步驟序列、outcome gating）已於 2026-04-28 移至 **adoption layer（XSPEC-095）**。此處保留 8 大審查類別定義、BLOCKING/IMPORTANT/SUGGESTION/QUESTION/NOTE 前綴語意供參考。
+> Lifecycle orchestration has moved to the adoption layer (XSPEC-095); this document is **reference-only**. For enforced execution, use your adoption layer's toolchain.
+
 Perform systematic code review using standardized checklists and comment prefixes.
 
 執行系統性的程式碼審查，使用標準化的檢查清單和評論前綴。
 
 ## Workflow | 工作流程
 
+> 📖 以下為流程的**參考結構**，非可執行步驟；自動化／強制執行請用 adoption layer 工具鏈（XSPEC-095）。
+> The steps below are a **reference structure**, not an executable workflow.
+
 1. **Identify changes** - Get diff of files to review via `git diff` or `git show`
 2. **Apply checklist** - Check each review category systematically
 3. **Generate report** - Output findings with standard comment prefixes
@@ -49,16 +56,16 @@ Perform systematic code review using standardized checklists and comment prefixe
 
 ## Usage | 使用方式
 
-- `/review` - Review all changes in current branch
-- `/review src/auth.js` - Review specific file
-- `/review feature/login` - Review specific branch
+- `/code-review` - Review all changes in current branch
+- `/code-review src/auth.js` - Review specific file
+- `/code-review feature/login` - Review specific branch
 
 ## Next Steps Guidance | 下一步引導
 
-After `/review` completes, the AI assistant should suggest:
+After `/code-review` completes, the AI assistant should suggest:
 
 > **程式碼審查完成。建議下一步 / Code review complete. Suggested next steps:**
-> - 有 ❗ BLOCKING 項目 → 修復後重新執行 `/review` — Fix issues then re-run `/review`
+> - 有 ❗ BLOCKING 項目 → 修復後重新執行 `/code-review` — Fix issues then re-run `/code-review`
 > - 全部通過 → 執行 `/checkin` 品質關卡 ⭐ **Recommended / 推薦** — All passed → Run `/checkin` quality gates
 > - 僅有 💡 SUGGESTION → 執行 `/commit` 提交變更 — Only suggestions → Run `/commit`
 > - 審查中發現規範不實用或缺失 → 執行 `/audit --report` 回報 — Found impractical or missing standards → Run `/audit --report`
@@ -71,6 +78,6 @@ After `/review` completes, the AI assistant should suggest:
 
 ## AI Agent Behavior | AI 代理行為
 
-> 完整的 AI 行為定義請參閱對應的命令文件：[`/review`](../commands/review.md#ai-agent-behavior--ai-代理行為)
+> 完整的 AI 行為定義請參閱對應的命令文件：[`/code-review`](../commands/code-review.md#ai-agent-behavior--ai-代理行為)
 >
-> For complete AI agent behavior definition, see the corresponding command file: [`/review`](../commands/review.md#ai-agent-behavior--ai-代理行為)
+> For complete AI agent behavior definition, see the corresponding command file: [`/code-review`](../commands/code-review.md#ai-agent-behavior--ai-代理行為)

package/bundled/skills/commands/COMMAND-FAMILY-OVERVIEW.md

@@ -301,8 +301,8 @@ Is this a new feature with clear requirements?
 
 ## Related Documentation | 相關文件
 
-- [Forward Derivation SKILL](../forward-derivation/SKILL.md) - Detailed derivation workflow
-- [Integrated Flow Guide](../methodology-system/integrated-flow.md) - Complete methodology integration
+- [Forward Derivation SKILL](../spec-derivation/SKILL.md) - Detailed derivation workflow
+- [Integrated Flow Guide](../dev-methodology/integrated-flow.md) - Complete methodology integration
 - [Individual Command Files](./README.md) - All command references
 
 ---