universal-dev-standards 5.7.2 → 5.8.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/bundled/ai/standards/acceptance-criteria-traceability.ai.yaml +49 -3
- package/bundled/ai/standards/behavior-snapshot.ai.yaml +168 -0
- package/bundled/ai/standards/feature-manifest-standard.ai.yaml +146 -0
- package/bundled/core/acceptance-criteria-traceability.md +41 -6
- package/bundled/core/behavior-snapshot.md +239 -0
- package/bundled/core/feature-manifest-standard.md +213 -0
- package/bundled/locales/zh-CN/CHANGELOG.md +17 -3
- package/bundled/locales/zh-CN/README.md +1 -1
- package/bundled/locales/zh-TW/CHANGELOG.md +17 -3
- package/bundled/locales/zh-TW/README.md +1 -1
- package/bundled/skills/brainstorm-assistant/SKILL.md +183 -30
- package/bundled/skills/brainstorm-assistant/guide.md +607 -117
- package/package.json +1 -1
- package/src/commands/update.js +12 -16
- package/standards-registry.json +29 -5
|
@@ -10,17 +10,71 @@ argument-hint: "[problem or feature idea | 問題或功能構想]"
|
|
|
10
10
|
|
|
11
11
|
> **Language**: English | [繁體中文](../../locales/zh-TW/skills/brainstorm-assistant/SKILL.md)
|
|
12
12
|
|
|
13
|
-
Structured ideation before specification writing. Transform vague ideas into actionable feature proposals through guided brainstorming.
|
|
13
|
+
Structured ideation before specification writing. Transform vague ideas into actionable feature proposals through guided brainstorming — grounded in cognitive science research.
|
|
14
14
|
|
|
15
|
-
|
|
15
|
+
在撰寫規格前進行結構化發想。以認知科學研究為基礎,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。
|
|
16
|
+
|
|
17
|
+
> **Implements**: XSPEC-196 brainstorm-assistant v2 — Cognitive Science Upgrade
|
|
18
|
+
|
|
19
|
+
## Mode Selection | 使用前先選模式
|
|
20
|
+
|
|
21
|
+
Apply these **objective triggers** before starting. Default is full v2.0 — routing rules are shortcuts to skip phases, not barriers to add.
|
|
22
|
+
|
|
23
|
+
使用前套用以下**客觀觸發條件**。預設為完整 v2.0,路由規則是跳過階段的快捷鍵,而非額外障礙。
|
|
24
|
+
|
|
25
|
+
| Condition | Recommended Mode | Command |
|
|
26
|
+
|-----------|-----------------|---------|
|
|
27
|
+
| Problem description < 20 words **or** topic feels vague | Full v2.0 (default) | `/brainstorm [topic]` |
|
|
28
|
+
| Strategic question (career, architecture, business model) | Full v2.0 with rebuttal | `/brainstorm [topic]` |
|
|
29
|
+
| Creative-only (naming, tagline, marketing copy) | Lite — skip rebuttal | `/brainstorm --no-rebuttal [topic]` |
|
|
30
|
+
| Time-constrained or execution-type (write code, fix copy) | Quick mode | `/brainstorm --quick [topic]` |
|
|
31
|
+
| Already have an SDD spec for this topic | Skip pre-flight | `/brainstorm --skip-preflight [topic]` |
|
|
32
|
+
|
|
33
|
+
> **Rule of thumb:** If you are unsure which row applies, use full v2.0. The cognitive overhead of deciding is higher than just running the full flow.
|
|
34
|
+
>
|
|
35
|
+
> **判斷原則:** 不確定適用哪一行時,直接用完整 v2.0。判斷本身的認知成本高於直接跑完整流程。
|
|
16
36
|
|
|
17
37
|
## Workflow | 工作流程
|
|
18
38
|
|
|
19
39
|
```
|
|
20
|
-
FRAME ──► DIVERGE ──► CONVERGE ──► OUTPUT
|
|
21
|
-
定義問題
|
|
40
|
+
[Mode Selection] ──► PRE-FLIGHT ──► FRAME ──► DIVERGE ──► CONVERGE ──► OUTPUT
|
|
41
|
+
客觀路由 防止錨定 定義問題 發散思考 反駁收斂 輸出提案
|
|
22
42
|
```
|
|
23
43
|
|
|
44
|
+
---
|
|
45
|
+
|
|
46
|
+
### Phase 0: PRE-FLIGHT | 防止 AI 錨定(新增)
|
|
47
|
+
|
|
48
|
+
**Why this phase exists:** Research (Nominal Group Technique) shows that seeing
|
|
49
|
+
AI output first anchors your thinking to the AI's training-data centre. Writing
|
|
50
|
+
your own ideas *before* the AI generates anything consistently produces more
|
|
51
|
+
diverse and higher-quality results.
|
|
52
|
+
|
|
53
|
+
**本階段存在的原因:** Nominal Group Technique 研究顯示,先看到 AI 輸出會將思考錨定在
|
|
54
|
+
AI 的訓練資料中心。在 AI 生成任何內容之前先寫下自己的想法,能持續產出更多樣、更高品質的結果。
|
|
55
|
+
|
|
56
|
+
**Before the AI generates any content**, the user completes three items:
|
|
57
|
+
|
|
58
|
+
在 AI 生成任何內容之前,使用者完成三件事:
|
|
59
|
+
|
|
60
|
+
| Item | Prompt | 說明 |
|
|
61
|
+
|------|--------|------|
|
|
62
|
+
| 1 | One-sentence problem description | 一句話描述問題 |
|
|
63
|
+
| 2 | Three initial ideas (any format, any quality) | 3 個初始想法(任意形式、不限品質) |
|
|
64
|
+
| 3 | "Solution types I do NOT want" (N/A allowed) | 「我最不想要的解法類型」(可填 N/A) |
|
|
65
|
+
|
|
66
|
+
**After the user submits**, the AI reads all three inputs and proceeds to FRAME.
|
|
67
|
+
The AI's first DIVERGE batch MUST be labelled "Exploring directions you haven't
|
|
68
|
+
mentioned" and MUST NOT duplicate any of the user's three submitted ideas.
|
|
69
|
+
|
|
70
|
+
**使用者提交後**,AI 讀取全部三項輸入再進入 FRAME。AI 的第一批 DIVERGE 輸出必須標示
|
|
71
|
+
「探索你尚未提及的方向」,且不得重複使用者已提交的任何想法。
|
|
72
|
+
|
|
73
|
+
**Flag:** `--skip-preflight` bypasses this phase with a one-line warning:
|
|
74
|
+
`⚠ Skipping Pre-flight may cause AI anchoring`
|
|
75
|
+
|
|
76
|
+
---
|
|
77
|
+
|
|
24
78
|
### Phase 1: FRAME | 定義問題
|
|
25
79
|
|
|
26
80
|
Define the problem space clearly before generating ideas.
|
|
@@ -29,16 +83,46 @@ Define the problem space clearly before generating ideas.
|
|
|
29
83
|
|
|
30
84
|
| Step | Action | 步驟 |
|
|
31
85
|
|------|--------|------|
|
|
32
|
-
| 1 | Clarify the problem with 5 Whys | 用 5 Whys
|
|
86
|
+
| 1 | Clarify the problem with 5 Whys | 用 5 Whys 釐清問題根因 |
|
|
33
87
|
| 2 | Reframe as "How Might We" (HMW) questions | 重構為 HMW 問題 |
|
|
34
|
-
| 3 | Identify stakeholders and constraints |
|
|
35
|
-
| 4 | Gather context from codebase (if applicable) |
|
|
88
|
+
| 3 | Identify stakeholders and constraints | 識別利害關係人與限制條件 |
|
|
89
|
+
| 4 | Gather context from codebase (if applicable) | 從程式碼庫蒐集脈絡(如適用) |
|
|
90
|
+
|
|
91
|
+
---
|
|
36
92
|
|
|
37
93
|
### Phase 2: DIVERGE | 發散思考
|
|
38
94
|
|
|
39
|
-
Generate
|
|
95
|
+
Generate ideas in **two semantic batches**. A minimum of **10 ideas** is
|
|
96
|
+
required before CONVERGE unlocks. The "Enter CONVERGE" option is hidden until
|
|
97
|
+
this gate is met.
|
|
98
|
+
|
|
99
|
+
分兩個語義批次產生想法。CONVERGE 解鎖前需累積最少 **10 個想法**,達標前「進入 CONVERGE」
|
|
100
|
+
選項不顯示。
|
|
101
|
+
|
|
102
|
+
#### Batch 1 — Intuition Batch (ideas 1–5) | 直覺批
|
|
103
|
+
|
|
104
|
+
Generate fast, unfiltered ideas. Label this batch "Intuition Batch — fast,
|
|
105
|
+
unfiltered". Upon completing idea 5, display:
|
|
106
|
+
`✓ Intuition batch complete`
|
|
107
|
+
|
|
108
|
+
快速、不過濾地產生想法。完成第 5 個後顯示:`✓ 直覺批完成`
|
|
109
|
+
|
|
110
|
+
#### Batch 2 — Extension Batch (ideas 6–10) | 延伸批
|
|
40
111
|
|
|
41
|
-
|
|
112
|
+
Display before starting: "Extension Batch: ideas must cross the semantic
|
|
113
|
+
boundary of Batch 1". If a proposed idea shares a theme type with any Batch 1
|
|
114
|
+
idea, flag: `⚠ Semantic overlap — try a different direction` (non-blocking).
|
|
115
|
+
|
|
116
|
+
開始前顯示延伸批提示。若想法主題與第一批重複,顯示警告(非阻擋)。
|
|
117
|
+
|
|
118
|
+
#### Gate | 門檻
|
|
119
|
+
|
|
120
|
+
- Below 10 ideas: show `Continue diverging (N/10)`, do not offer CONVERGE
|
|
121
|
+
- At 10+ ideas: unlock CONVERGE; no upper limit on continuing
|
|
122
|
+
|
|
123
|
+
少於 10 個:顯示「繼續發散(N/10)」;10 個以上:解鎖 CONVERGE,可繼續無上限。
|
|
124
|
+
|
|
125
|
+
#### Techniques | 技法
|
|
42
126
|
|
|
43
127
|
| Technique | When to Use | 使用時機 |
|
|
44
128
|
|-----------|-------------|----------|
|
|
@@ -46,11 +130,15 @@ Generate as many ideas as possible without judgment.
|
|
|
46
130
|
| **SCAMPER** | Improving existing features | 改善現有功能 |
|
|
47
131
|
| **Six Thinking Hats** | Need multiple perspectives | 需要多角度思考 |
|
|
48
132
|
|
|
49
|
-
|
|
133
|
+
---
|
|
134
|
+
|
|
135
|
+
### Phase 3: CONVERGE | 反駁收斂
|
|
50
136
|
|
|
51
|
-
|
|
137
|
+
Two sub-steps: scoring, then the **Rebuttal Round**.
|
|
52
138
|
|
|
53
|
-
|
|
139
|
+
兩個子步驟:評分,然後進行**反駁輪**。
|
|
140
|
+
|
|
141
|
+
#### Step 3a: Scoring | 評分
|
|
54
142
|
|
|
55
143
|
| Criterion | Weight | 評估標準 |
|
|
56
144
|
|-----------|--------|----------|
|
|
@@ -59,21 +147,48 @@ Evaluate and prioritize ideas using structured criteria.
|
|
|
59
147
|
| Effort | 20% | 實作成本 |
|
|
60
148
|
| Alignment | 20% | 目標一致性 |
|
|
61
149
|
|
|
150
|
+
#### Step 3b: Rebuttal Round | 反駁輪(新增)
|
|
151
|
+
|
|
152
|
+
**Why this step exists:** Nemeth (1995) showed that groups allowed to debate
|
|
153
|
+
produce more and better ideas than groups following "no-criticism" rules.
|
|
154
|
+
Challenging the top ideas before committing prevents selecting ideas that merely
|
|
155
|
+
*sound* good under positive-only evaluation.
|
|
156
|
+
|
|
157
|
+
**本步驟存在的原因:** Nemeth (1995) 證明允許辯論的小組比遵守「禁止批評」規則的小組
|
|
158
|
+
產出更多更好的想法。在確定前挑戰最佳想法,能防止選出只是在正向評估下「聽起來不錯」的方案。
|
|
159
|
+
|
|
160
|
+
For each of the **top 3 scored ideas**, the AI presents **2 specific
|
|
161
|
+
counterarguments** in the format: "This idea will fail in [context] because
|
|
162
|
+
[reason]." Vague concerns ("this might be hard") are not acceptable.
|
|
163
|
+
|
|
164
|
+
對前三名想法,AI 各提出 **2 個具體反對理由**,格式為:「在 [情境] 下,此想法會失敗,
|
|
165
|
+
因為 [原因]。」模糊的顧慮(「這可能有點困難」)不可接受。
|
|
166
|
+
|
|
167
|
+
The user **must** respond to each counterargument before the session advances:
|
|
168
|
+
|
|
169
|
+
使用者必須對每個反對理由給出回應才能繼續:
|
|
170
|
+
|
|
171
|
+
| Option | Action | 說明 |
|
|
172
|
+
|--------|--------|------|
|
|
173
|
+
| (a) | Accept criticism → provide modified version | 接受批評 → 提供修改版本 |
|
|
174
|
+
| (b) | Disagree → provide reason to retain | 不同意 → 說明保留理由 |
|
|
175
|
+
| (c) | Criticism valid → remove from ranking | 批評成立 → 從排名移除 |
|
|
176
|
+
|
|
177
|
+
**Flag:** `--no-rebuttal` skips this step; report section marked
|
|
178
|
+
"Rebuttal: skipped".
|
|
179
|
+
|
|
180
|
+
---
|
|
181
|
+
|
|
62
182
|
### Phase 4: OUTPUT | 輸出提案
|
|
63
183
|
|
|
64
184
|
Produce a Brainstorm Report ready for `/requirement` or `/sdd`.
|
|
65
185
|
|
|
66
186
|
產生可直接對接 `/requirement` 或 `/sdd` 的腦力激盪報告。
|
|
67
187
|
|
|
68
|
-
|
|
188
|
+
Each idea that passed the rebuttal round is marked `✓ Passed rebuttal` with a
|
|
189
|
+
one-line summary of the user's response.
|
|
69
190
|
|
|
70
|
-
|
|
71
|
-
|-----------|---------|-------|------|
|
|
72
|
-
| **5 Whys** | Root cause analysis | Ask "Why?" 5 times | 根因分析 |
|
|
73
|
-
| **HMW** | Problem reframing | "How might we [verb] [outcome]?" | 問題重構 |
|
|
74
|
-
| **SCAMPER** | Idea modification | 7 prompts: Substitute, Combine, Adapt, Modify, Put to other use, Eliminate, Reverse | 創意改造 |
|
|
75
|
-
| **Six Thinking Hats** | Multi-perspective | 6 modes: Facts, Emotions, Risks, Benefits, Creativity, Process | 多角度思考 |
|
|
76
|
-
| **Dot Voting** | Quick prioritization | Each participant gets 3 votes | 快速排序 |
|
|
191
|
+
通過反駁輪的想法標記 `✓ Passed rebuttal` 及使用者回應摘要一行。
|
|
77
192
|
|
|
78
193
|
## Output Format | 輸出格式
|
|
79
194
|
|
|
@@ -89,15 +204,15 @@ Produce a Brainstorm Report ready for `/requirement` or `/sdd`.
|
|
|
89
204
|
3. How might we ...?
|
|
90
205
|
|
|
91
206
|
## Ideas Generated
|
|
92
|
-
| # | Idea | Source Technique | Feasibility | Impact | Score |
|
|
93
|
-
|
|
94
|
-
| 1 | ... | SCAMPER | 4/5 | 5/5 | 4.3 |
|
|
95
|
-
| 2 | ... | HMW | 3/5 | 4/5 | 3.5 |
|
|
207
|
+
| # | Idea | Batch | Source Technique | Feasibility | Impact | Score |
|
|
208
|
+
|---|------|-------|-----------------|-------------|--------|-------|
|
|
209
|
+
| 1 | ... | B1 | SCAMPER | 4/5 | 5/5 | 4.3 |
|
|
210
|
+
| 2 | ... | B2 | HMW | 3/5 | 4/5 | 3.5 |
|
|
96
211
|
|
|
97
212
|
## Top 3 Recommendations
|
|
98
|
-
1. **[Idea Name]**
|
|
99
|
-
2. **[Idea Name]**
|
|
100
|
-
3. **[Idea Name]**
|
|
213
|
+
1. **[Idea Name]** ✓ Passed rebuttal — [Why recommended] — [User rebuttal response summary]
|
|
214
|
+
2. **[Idea Name]** ✓ Passed rebuttal — [Why recommended] — [User rebuttal response summary]
|
|
215
|
+
3. **[Idea Name]** ✓ Passed rebuttal — [Why recommended] — [User rebuttal response summary]
|
|
101
216
|
|
|
102
217
|
## Next Steps
|
|
103
218
|
- [ ] Proceed to `/requirement` with top idea
|
|
@@ -105,11 +220,49 @@ Produce a Brainstorm Report ready for `/requirement` or `/sdd`.
|
|
|
105
220
|
- [ ] Need further exploration of idea #N
|
|
106
221
|
```
|
|
107
222
|
|
|
223
|
+
## Technique Quick Reference | 技法速覽
|
|
224
|
+
|
|
225
|
+
| Technique | Purpose | Steps | 用途 |
|
|
226
|
+
|-----------|---------|-------|------|
|
|
227
|
+
| **5 Whys** | Root cause analysis | Ask "Why?" 5 times | 根因分析 |
|
|
228
|
+
| **HMW** | Problem reframing | "How might we [verb] [outcome]?" | 問題重構 |
|
|
229
|
+
| **SCAMPER** | Idea modification | 7 prompts: Substitute, Combine, Adapt, Modify, Put to other use, Eliminate, Reverse | 創意改造 |
|
|
230
|
+
| **Six Thinking Hats** | Multi-perspective | 6 modes: Facts, Emotions, Risks, Benefits, Creativity, Process | 多角度思考 |
|
|
231
|
+
| **Dot Voting** | Quick prioritization | Each participant gets 3 votes | 快速排序 |
|
|
232
|
+
|
|
233
|
+
## Session Self-Evaluation | 工作階段自評
|
|
234
|
+
|
|
235
|
+
After each session, record three metrics (1–5 scale) to track improvement over time.
|
|
236
|
+
|
|
237
|
+
每次工作階段結束後,記錄三個指標(1–5 分),追蹤長期改善趨勢。
|
|
238
|
+
|
|
239
|
+
| Metric | Question | 指標 |
|
|
240
|
+
|--------|----------|------|
|
|
241
|
+
| **Adoption Rate** | How many of today's ideas will I actually use? | 採用率:今天的想法有幾個會實際使用? |
|
|
242
|
+
| **Diversity** | Were the Extension Batch ideas noticeably different from Intuition Batch? | 多樣性:延伸批是否與直覺批明顯不同? |
|
|
243
|
+
| **Cognitive Load** | How mentally taxing was this session? (5 = effortless, 1 = exhausting) | 認知負擔:這次工作階段的心智消耗程度? |
|
|
244
|
+
|
|
245
|
+
Collect 3 sessions before drawing conclusions. See [guide.md](./guide.md) for the full A/B experiment protocol.
|
|
246
|
+
|
|
247
|
+
收集 3 次工作階段資料再下結論。完整 A/B 實驗協議見 [guide.md](./guide.md)。
|
|
248
|
+
|
|
249
|
+
## Flags | 旗標
|
|
250
|
+
|
|
251
|
+
| Flag | Effect | 說明 |
|
|
252
|
+
|------|--------|------|
|
|
253
|
+
| `--skip-preflight` | Skip Phase 0 with warning | 跳過 Phase 0,顯示錨定警告 |
|
|
254
|
+
| `--no-rebuttal` | Skip rebuttal round in CONVERGE | 跳過反駁輪,報告標注 skipped |
|
|
255
|
+
| `--quick` | 3-idea fast mode; gates and rebuttal exempt | 快速 3 想法模式;門檻與反駁均豁免 |
|
|
256
|
+
| `--technique scamper` | Force SCAMPER as primary technique | 強制使用 SCAMPER |
|
|
257
|
+
|
|
108
258
|
## Usage | 使用方式
|
|
109
259
|
|
|
110
|
-
- `/brainstorm`
|
|
111
|
-
- `/brainstorm "user retention"`
|
|
112
|
-
- `/brainstorm --
|
|
260
|
+
- `/brainstorm` — Start interactive brainstorming session
|
|
261
|
+
- `/brainstorm "user retention"` — Brainstorm around a specific topic
|
|
262
|
+
- `/brainstorm --quick "reduce checkout friction"` — Fast 3-idea mode
|
|
263
|
+
- `/brainstorm --skip-preflight "feature X"` — Skip Pre-flight (with warning)
|
|
264
|
+
- `/brainstorm --no-rebuttal "topic"` — Skip rebuttal round
|
|
265
|
+
- `/brainstorm --technique scamper` — Use a specific technique
|
|
113
266
|
|
|
114
267
|
## Next Steps Guidance | 下一步引導
|
|
115
268
|
|