sources.json
7.63 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
{
"task_id": "T-mt5ufepi",
"target_agent": "workgraph-planning-methodology-trial",
"generated_by": "innovator-hunt",
"version": "v1",
"blocks": {
"A_external_methodology": [
{
"topic": "Cynefin framework",
"claim": "五域（Clear/Complicated/Complex/Chaotic/Disorder）按因果可辨识程度分类，不同域需要不同应对策略（sense-categorize-respond / sense-analyse-respond / probe-sense-respond / act-sense-respond）",
"sources": [
"https://en.wikipedia.org/wiki/Cynefin_framework",
"https://untools.co/cynefin-framework/",
"https://modelthinkers.com/mental-model/cynefin-framework"
],
"verified": true
},
{
"topic": "T-shirt sizing / Story Points",
"claim": "T-shirt sizing 用相对档位（XS-XL）代替绝对工时估算，适合早期粗筛，再映射到具体 Story Point 数值区间做细粒度规划",
"sources": [
"https://asana.com/resources/t-shirt-sizing",
"https://teachingagile.com/scrum/psm-1/scrum-planning-estimation/estimation-techniques/t-shirt-sizing"
],
"verified": true
},
{
"topic": "INVEST principles",
"claim": "好用户故事应 Independent/Negotiable/Valuable/Estimable/Small/Testable；够小的判据之一是 6-10 个故事应能在一个 sprint 内完成",
"sources": [
"https://climbtheladder.com/what-makes-a-good-user-story-invest-acceptance-criteria/"
],
"verified": true
},
{
"topic": "SPIDR splitting patterns",
"claim": "Mike Cohn 提出的 5 种拆分手法：Spike/Path/Interface/Data/Rules；与 INVEST 是互补关系而非替代",
"sources": [
"https://www.mountaingoatsoftware.com/agile/five-simple-but-powerful-ways-to-split-user-stories",
"https://blogs.itemis.com/en/spidr-five-simple-techniques-for-a-perfectly-split-user-story"
],
"verified": true
},
{
"topic": "YAGNI / Lean",
"claim": "不要在真正需要前添加功能，做能work的最简单方案；核心论据是维护成本（理解/测试/调试/更新）",
"sources": [
"https://lawsofsoftwareengineering.com/laws/yagni/"
],
"verified": true
},
{
"topic": "YAGNI 矫枉过正风险",
"claim": "YAGNI 与架构规划存在张力，'不能靠以后加楼层来盖摩天大楼'，YAGNI适用于功能/抽象层而非地基级架构决策",
"sources": [
"https://news.ycombinator.com/item?id=33235887",
"https://lawsofsoftwareengineering.com/laws/yagni/"
],
"verified": true
},
{
"topic": "'64% 功能很少或从未被使用' 统计数字",
"claim": "多篇二手博客引用此数字支持YAGNI，但未找到可验证的一手研究链接",
"sources": [],
"verified": false,
"note": "未核实——反复出现在二手来源中，未找到原始研究出处"
}
],
"B_internal_authoritative": [
{
"topic": "workgraph-planning agent.md 全文",
"file": "system/agents/workgraph-planning/agent.md",
"sections_quoted": [
"Step 2 读候选骨架 (L75-86)",
"Step 4 规划·裁剪授权原句 (L117-148)",
"Step 5 落盘规划记录·七节结构 (L163-209)",
"约束一·五项校验原文 (L405-424)"
],
"verified": true
},
{
"topic": "planner.json 骨架定义",
"file": "system/work-graph-templates/planner.json",
"claim": "5节点 brd→prd→sdd→benchmark→validator；benchmark→validator 边 needs:[brd,prd,sdd]",
"verified": true
},
{
"topic": "executor.json 骨架定义",
"file": "system/work-graph-templates/executor.json",
"claim": "3节点 impl→test→validator，含占位符 {{orgNode}}/{{implAgent}}/{{testAgent}}/{{validatorAgent}}",
"verified": true
},
{
"topic": "executor-fanout.json 骨架定义",
"file": "system/work-graph-templates/executor-fanout.json",
"claim": "8节点 split→(impl/test)×2→merge→integration→validator",
"verified": true
},
{
"topic": "validator.json 骨架定义",
"file": "system/work-graph-templates/validator.json",
"claim": "3节点 validator-triager→benchmark-validator→testdata-validator，两条forward边只写needs不写produces",
"verified": true
},
{
"topic": "work-graph-templates README",
"file": "system/work-graph-templates/README.md",
"claim": "四张骨架共同形状'做→自检→本层验收'；砍掉 org-planner-brd 会触发 rule=data-contract（因 benchmark→validator 边 needs 里的 brd 找不到产出方）",
"verified": true
}
],
"C_historical_cases": [
{
"case_id": "p-v82tpc/T-mt2r6fz0-125b",
"file": "data/companies/e2e-demo/projects/p-v82tpc/workgraph-plan-T-mt2r6fz0-125b.md",
"host_org_node": "org-validator",
"skeleton_used": "validator.json",
"pruned": false,
"note": "完整3节点，与骨架完全一致",
"verified": true
},
{
"case_id": "p-pyjha7/T-mt45nzvc-2235",
"file": "data/companies/e2e-demo/projects/p-pyjha7/workgraph-plan-T-mt45nzvc-2235.md",
"host_org_node": "org-validator",
"skeleton_used": "validator.json",
"pruned": false,
"note": "完整3节点；testdata-validator 保留即使 goal-test-data/ 不存在，靠 required:false 逃生",
"verified": true
},
{
"case_id": "p-iyshuc/T-mt5d3mi9-57cf (v1)",
"file": "data/companies/e2e-demo/projects/p-iyshuc/workgraph-plan-T-mt5d3mi9-57cf.md",
"host_org_node": "org-coder",
"skeleton_used": "executor.json",
"pruned": false,
"note": "完整3节点，唯一一处显式排除 executor-fanout.json 并给出理由（三卖点区块同属一次实现，非独立并行任务）",
"verified": true
},
{
"case_id": "p-iyshuc/T-mt5d3mi9-57cf (retry2)",
"file": "data/companies/e2e-demo/projects/p-iyshuc/workgraph-plan-T-mt5d3mi9-57cf-retry2.md",
"host_org_node": "org-coder",
"skeleton_used": "executor.json",
"pruned": false,
"note": "infra_flake归因，复用上一版图未重新设计，落地了 agent.md Step4 重规划三选一表格第三行的判据",
"verified": true
},
{
"case_id": "p-v1mpcl/T-mt1qx0c4-95bf (补充，非任务原指定)",
"file": "data/companies/e2e-demo/projects/p-v1mpcl/workgraph-plan-T-mt1qx0c4-95bf.md",
"host_org_node": "org-planner",
"skeleton_used": "planner.json",
"pruned": false,
"note": "全公司唯一一个org-planner真实案例，任务特征与Step4裁剪例句高度吻合（读文档产出规格化静态网站），但完整保留5节点包括BRD，是本次调研发现的唯一'与裁剪例句直接可比但未裁剪'的反例",
"verified": true
}
]
},
"summary_finding": "全公司当前可查到的5份work graph规划记录（覆盖4个真实任务）中，裁剪发生次数为0；骨架选型100%由宿主org node类型确定性决定，未见任务复杂度参与选型判断；唯一一次'骨架间比较并排除'的记录出现在p-iyshuc案例中（排除executor-fanout.json）；p-v1mpcl(C4)是唯一与agent.md Step4裁剪例句高度可比的真实样本，结果为完整保留而非裁剪，需要后续人工核实该BRD节点产出是否属实需要（本次未核实）。"
}