@krr2020/taskflow-mcp-server 0.1.0-beta.1 → 0.1.0-beta.3

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.
@@ -0,0 +1,171 @@
1
+ # 🔨 TASK GENERATOR
2
+
3
+ **Purpose:** Break PRDs into Features → Stories → Tasks (JSON-based)
4
+
5
+ ---
6
+
7
+ ## Directory Structure
8
+
9
+ ```
10
+ tasks/
11
+ ├── project-index.json # Root index
12
+ └── F[N]-[feature-name]/ # Feature folder
13
+ ├── F[N]-[feature-name].json # Stories & tasks list
14
+ └── S[N].[M]-[story-name]/ # Story folder
15
+ └── T[N].[M].[K]-[task-title].json # Task file
16
+ ```
17
+
18
+ ---
19
+
20
+ ## ID Conventions
21
+
22
+ | Level | Format | Example |
23
+ |-------|--------|---------|
24
+ | Feature | `N` | `1`, `2` |
25
+ | Story | `N.M` | `1.1`, `1.2` |
26
+ | Task | `N.M.K` | `1.1.0`, `1.1.1` |
27
+
28
+ **No leading zeros.** `1.1.0` ✓ `1.01.0` ✗
29
+
30
+ ---
31
+
32
+ ## Generation Workflow
33
+
34
+ 1. Parse PRD → Identify Features & Stories
35
+ 2. Create folder structure
36
+ 3. Generate `project-index.json`
37
+ 4. Generate `F[N]-[feature-name].json` per feature
38
+ 5. Generate task JSON files
39
+
40
+ ---
41
+
42
+ ## Templates
43
+
44
+ ### project-index.json
45
+ ```json
46
+ {
47
+ "project": "Project Name",
48
+ "features": [
49
+ {
50
+ "id": "1",
51
+ "title": "Feature Title",
52
+ "status": "not-started",
53
+ "path": "F1-feature-name"
54
+ }
55
+ ]
56
+ }
57
+ ```
58
+
59
+ ### F[N]-[feature-name].json
60
+ ```json
61
+ {
62
+ "id": "1",
63
+ "title": "Feature Title",
64
+ "status": "not-started",
65
+ "stories": [
66
+ {
67
+ "id": "1.1",
68
+ "title": "Story Title",
69
+ "status": "not-started",
70
+ "tasks": [
71
+ {
72
+ "id": "1.1.0",
73
+ "title": "Task Title",
74
+ "status": "not-started",
75
+ "dependencies": []
76
+ }
77
+ ]
78
+ }
79
+ ]
80
+ }
81
+ ```
82
+
83
+ ### Task File (T1.1.0-setup-database-schema.json)
84
+ ```json
85
+ {
86
+ "id": "1.1.0",
87
+ "title": "Setup Database Schema",
88
+ "description": "Create migration files and initial schema definitions.",
89
+ "status": "not-started",
90
+ "skill": "backend",
91
+ "subtasks": [
92
+ {
93
+ "id": "1",
94
+ "description": "Create migration file",
95
+ "status": "pending"
96
+ },
97
+ {
98
+ "id": "2",
99
+ "description": "Define entity in schema",
100
+ "status": "pending"
101
+ }
102
+ ],
103
+ "context": [
104
+ "src/database/schema.ts",
105
+ "src/database/schema2.ts"
106
+ ]
107
+ }
108
+ ```
109
+
110
+ ### Task Status Values
111
+
112
+ | Status | Description |
113
+ |--------|-------------|
114
+ | `not-started` | Task not yet begun |
115
+ | `setup` | Reading context and requirements |
116
+ | `implementing` | Writing code |
117
+ | `verifying` | Self-review |
118
+ | `validating` | Running automated checks |
119
+ | `committing` | Ready to commit and push |
120
+ | `completed` | Task finished |
121
+ | `blocked` | Cannot proceed |
122
+ | `on-hold` | Paused |
123
+
124
+ ### Skill Values
125
+
126
+ | Skill | Use When |
127
+ |-------|----------|
128
+ | `development` | General development tasks, refactoring, utilities |
129
+ | `backend` | API, database, server logic, handlers, repositories |
130
+ | `frontend` | UI components, pages, forms, client-side logic |
131
+ | `fullstack` | End-to-end features crossing frontend/backend |
132
+ | `devops` | Docker, K8s, CI/CD, deployment, infrastructure |
133
+ | `docs` | Documentation, README, guides |
134
+
135
+ **Rule:** Tasks MUST have single responsibility. Avoid mixing `frontend` and `backend` in one task. Use `fullstack` only for true integration work where contract definition matters.
136
+
137
+ ---
138
+
139
+ ## Task Granularity & Scoping
140
+
141
+ **CRITICAL:** The AI must optimize for **MINIMAL** number of tasks.
142
+ - **Preference:** Fewer, larger tasks > Many small tasks.
143
+ - **Metric:** A functional feature should typically be **1-2 stories**, and each story **2-4 tasks**.
144
+
145
+ ### ✅ Good Task Design
146
+ - **Scope:** A task should complete a meaningful unit of work (e.g., "Implement entire API endpoint with tests" or "Create full UI page with components").
147
+ - **Size:** 30-60 minutes of coding work.
148
+ - **Files:** Touches multiple files (controller, service, types, tests).
149
+
150
+ ### ❌ Anti-Patterns (AVOID THESE)
151
+ - **One task per file:** ❌ `T1.1.0-create-controller` (Too small)
152
+ - **One task per function:** ❌ `T1.1.0-add-validation-function` (Too small)
153
+ - **Separating simple steps:** ❌ `T1.1.0-install-package` -> `T1.1.1-config-package` (Merge these!)
154
+
155
+ ### Example Grouping
156
+ **Instead of:**
157
+ 1. `Create migration`
158
+ 2. `Create model`
159
+ 3. `Create repo`
160
+ 4. `Create service`
161
+
162
+ **DO THIS:**
163
+ 1. `T1.1.0-implement-user-persistence` (Includes migration, model, repo, service)
164
+
165
+ ---
166
+
167
+ ## Rules
168
+
169
+ - **Sequential stories:** Complete all tasks in one story before starting another.
170
+ - **Branch per story:** Each story gets `story/S[N].[M]-[story-name]` branch.
171
+ - **Kebab-case:** All folder/file names use kebab-case.