@torka/claude-workflows 0.11.0 → 0.13.1

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.
Files changed (40) hide show
  1. package/.claude-plugin/plugin.json +12 -0
  2. package/README.md +22 -5
  3. package/bmad-workflows/bmm/workflows/4-implementation/implement-epic-with-subagents/steps/step-01b-continue.md +9 -2
  4. package/bmad-workflows/bmm/workflows/4-implementation/implement-epic-with-subagents/steps/step-02-orchestrate.md +108 -2
  5. package/bmad-workflows/bmm/workflows/4-implementation/implement-epic-with-subagents/steps/step-03-complete.md +35 -1
  6. package/commands/deep-audit.md +389 -0
  7. package/commands/dev-story-backend.md +12 -11
  8. package/commands/dev-story-fullstack.md +6 -2
  9. package/commands/dev-story-ui.md +4 -4
  10. package/commands/github-pr-resolve.md +132 -24
  11. package/install.js +26 -4
  12. package/package.json +1 -1
  13. package/skills/deep-audit/INSPIRATIONS.md +26 -0
  14. package/skills/deep-audit/SKILL.md +253 -0
  15. package/skills/deep-audit/agents/api-contract-reviewer.md +38 -0
  16. package/skills/deep-audit/agents/architecture-and-complexity.md +48 -0
  17. package/skills/deep-audit/agents/code-health.md +51 -0
  18. package/skills/deep-audit/agents/data-layer-reviewer.md +39 -0
  19. package/skills/deep-audit/agents/performance-profiler.md +38 -0
  20. package/skills/deep-audit/agents/security-and-error-handling.md +52 -0
  21. package/skills/deep-audit/agents/seo-accessibility-auditor.md +49 -0
  22. package/skills/deep-audit/agents/test-coverage-analyst.md +37 -0
  23. package/skills/deep-audit/agents/type-design-analyzer.md +38 -0
  24. package/skills/deep-audit/templates/report-template.md +87 -0
  25. package/skills/designer-founder/SKILL.md +8 -7
  26. package/skills/designer-founder/steps/step-01-context.md +94 -45
  27. package/skills/designer-founder/steps/step-02-scope.md +6 -23
  28. package/skills/designer-founder/steps/step-03-design.md +29 -58
  29. package/skills/designer-founder/steps/step-04-artifacts.md +137 -113
  30. package/skills/designer-founder/steps/step-05-epic-linking.md +81 -53
  31. package/skills/designer-founder/steps/step-06-validate.md +181 -0
  32. package/skills/designer-founder/templates/component-strategy.md +4 -0
  33. package/skills/designer-founder/tools/magicpatterns.md +52 -19
  34. package/skills/designer-founder/tools/stitch.md +97 -67
  35. package/skills/product-architect/SKILL.md +308 -0
  36. package/skills/product-architect/agents/architect-agent.md +64 -0
  37. package/skills/product-architect/agents/pm-agent.md +65 -0
  38. package/skills/product-architect/references/escalation-guide.md +70 -0
  39. package/skills/product-architect/vt-preferences.md +44 -0
  40. package/uninstall.js +36 -0
@@ -2,10 +2,10 @@
2
2
 
3
3
  ## MANDATORY EXECUTION RULES (READ FIRST)
4
4
 
5
- - 🛑 NEVER assume project state without checking
6
- - 📖 CRITICAL: Read complete step file before taking action
7
- - ALWAYS treat this as collaborative discovery
8
- - 🎯 Goal: Understand where we are and what mode to use
5
+ - NEVER assume project state without checking
6
+ - CRITICAL: Read complete step file before taking action
7
+ - ALWAYS treat this as collaborative discovery
8
+ - Goal: Understand where we are and what mode to use
9
9
 
10
10
  ---
11
11
 
@@ -17,6 +17,34 @@ Establish context and select the appropriate workflow mode based on project stat
17
17
 
18
18
  ## TASK SEQUENCE
19
19
 
20
+ ### 0. Check for Returning Session
21
+
22
+ Before greeting, check for a previous session state file:
23
+
24
+ **Look for:** `{planning_artifacts}/ux-design/.designer-state.yaml`
25
+
26
+ **If found:**
27
+ ```
28
+ Welcome back, {user_name}! I see a previous design session:
29
+
30
+ Previous session:
31
+ - Tool: {tool_used}
32
+ - Theme: {theme_path_or_none}
33
+ - Mode: {mode}
34
+
35
+ [Y] Same settings - Jump to Step 2 (Scope)
36
+ [R] Review - Show full state, let me modify
37
+ [N] Start fresh - Run full setup
38
+ ```
39
+
40
+ - **Y**: Load previous state into working memory, skip to Step 2 (Scope)
41
+ - **R**: Show all saved state fields, let user confirm/modify each, then proceed
42
+ - **N**: Continue with normal Step 1 below
43
+
44
+ **If not found:** Continue with normal Step 1 below.
45
+
46
+ ---
47
+
20
48
  ### 1. Greet and Understand Intent
21
49
 
22
50
  Start with a brief greeting using `{user_name}` and ask what they'd like to design:
@@ -41,8 +69,8 @@ After user describes their intent, quickly assess:
41
69
  - Epic/story files related to user's intent?
42
70
 
43
71
  **Determine project phase:**
44
- - **Greenfield**: No relevant specs found Need to establish design direction
45
- - **Mid-project**: Specs exist Pull context, focus on specific scope
72
+ - **Greenfield**: No relevant specs found -> Need to establish design direction
73
+ - **Mid-project**: Specs exist -> Pull context, focus on specific scope
46
74
 
47
75
  **Report findings concisely:**
48
76
  ```
@@ -54,23 +82,58 @@ Project Context:
54
82
 
55
83
  ---
56
84
 
57
- ### 3. Check Available Tools
85
+ ### 3. Detect Theme
86
+
87
+ Search for theme files in the project. Themes follow a two-file convention:
88
+ - `{name}.prompt.md` -- Human-readable design rules (vibe, typography, components, animation)
89
+ - `{name}.tokens.json` -- Machine-readable HSL values for shadcn CSS variables
90
+
91
+ **Detection:** Search for `**/themes/*.prompt.md` or `**/themes/*.tokens.json` in:
92
+ 1. Project root
93
+ 2. One level up (`../`)
94
+
95
+ **If theme files found:**
96
+ ```
97
+ Theme detected: {theme_name}
98
+ - Prompt: {path to .prompt.md}
99
+ - Tokens: {path to .tokens.json}
100
+
101
+ Should I use these for design consistency? (Recommended)
102
+ [Y] Yes - Use this theme
103
+ [N] No - Continue without theme
104
+ ```
105
+
106
+ **If theme files NOT found:**
107
+ ```
108
+ No theme files detected in the project.
109
+
110
+ [P] Provide theme - Point me to your theme files
111
+ [S] Skip - Continue without a theme (tool defaults will be used)
112
+ ```
113
+
114
+ **If theme is from outside the project** (e.g., `../vt-design-studio/themes/`):
115
+ - Copy theme files to `{planning_artifacts}/themes/`
116
+ - Reference the local copy in all subsequent steps
117
+
118
+ ---
119
+
120
+ ### 4. Check Available Tools
58
121
 
59
122
  Detect and report tool availability:
60
123
 
61
124
  ```
62
125
  Available Tools:
63
- SuperDesign - HTML/CSS prototyping
64
- [✓/✗] MagicPatterns MCP - React component generation
65
- [✓/✗] shadcn MCP - Component search & install
66
- [✓/✗] Playwright MCP - Screenshot verification
126
+ [check] SuperDesign - HTML/CSS prototyping
127
+ [check] MagicPatterns MCP - React component generation
128
+ [check] Stitch MCP - Google AI design tool
129
+ [check] shadcn MCP - Component search & install
67
130
  ```
68
131
 
69
132
  Note: shadcn CLI (`npx shadcn@latest add`) is always available as fallback.
70
133
 
71
134
  ---
72
135
 
73
- ### 4. Present Mode Selection
136
+ ### 5. Present Mode Selection
74
137
 
75
138
  Based on context, recommend a mode and let user choose:
76
139
 
@@ -80,27 +143,27 @@ WORKFLOW MODE
80
143
  Based on your request, I recommend: [Quick Prototype / Production Flow]
81
144
 
82
145
  [Q] Quick Prototype
83
- Fast visual exploration
84
- Output: HTML prototype or wireframe
85
- Best for: Testing ideas, exploring directions
146
+ -> Fast visual exploration
147
+ -> Output: HTML prototype or wireframe
148
+ -> Best for: Testing ideas, exploring directions
86
149
 
87
150
  [P] Production Flow
88
- Full dev-ready artifacts
89
- Output: Component strategy, layouts, user journeys
90
- Best for: Features going into the product
151
+ -> Full dev-ready artifacts
152
+ -> Output: Component strategy, layouts, user journeys
153
+ -> Best for: Features going into the product
91
154
 
92
155
  Which mode? [Q/P]
93
156
  ```
94
157
 
95
158
  **Recommendation logic:**
96
- - User says "explore", "try", "prototype", "quick" Recommend Quick
97
- - User mentions specific epic/story, "build", "implement" Recommend Production
98
- - Greenfield project, first design Recommend Production (establish foundation)
99
- - Uncertain Ask user
159
+ - User says "explore", "try", "prototype", "quick" -> Recommend Quick
160
+ - User mentions specific epic/story, "build", "implement" -> Recommend Production
161
+ - Greenfield project, first design -> Recommend Production (establish foundation)
162
+ - Uncertain -> Ask user
100
163
 
101
164
  ---
102
165
 
103
- ### 5. Confirm and Route
166
+ ### 6. Confirm and Route
104
167
 
105
168
  Once user selects mode, confirm and route:
106
169
 
@@ -111,7 +174,7 @@ Quick Prototype mode selected.
111
174
  We'll skip detailed specs and focus on rapid visualization.
112
175
  Ready to move to design tool selection.
113
176
  ```
114
- Proceed to Step 3 (skip Step 2)
177
+ -> Proceed to Step 3 (skip Step 2)
115
178
 
116
179
  **If Production Flow:**
117
180
  ```
@@ -124,24 +187,7 @@ We'll create dev-ready artifacts including:
124
187
 
125
188
  Let's define the scope first.
126
189
  ```
127
- Proceed to Step 2
128
-
129
- ---
130
-
131
- ## COLLABORATION MENU
132
-
133
- After confirming mode, present:
134
-
135
- ```
136
- [A] Advanced - Deep dive into project context
137
- [P] Party Mode - Get multiple agent perspectives
138
- [C] Continue - Proceed to next step
139
- ```
140
-
141
- **Menu Handlers:**
142
- - **A**: Load `{project-root}/_bmad/core/workflows/advanced-elicitation/workflow.xml` for deeper discovery
143
- - **P**: Load `{project-root}/_bmad/core/workflows/party-mode/workflow.md` for multi-perspective input
144
- - **C**: Proceed to next step based on mode selection
190
+ -> Proceed to Step 2
145
191
 
146
192
  ---
147
193
 
@@ -157,8 +203,13 @@ related_artifacts: [list of relevant files found]
157
203
  tools_available:
158
204
  superdesign: true
159
205
  magicpatterns: [true/false]
206
+ stitch: [true/false]
160
207
  shadcn_mcp: [true/false]
161
- playwright: [true/false]
208
+ theme:
209
+ name: "{theme_name}"
210
+ prompt_file: "{path to .prompt.md}"
211
+ tokens_file: "{path to .tokens.json}"
212
+ is_local: true/false # whether files are in the project
162
213
  ```
163
214
 
164
215
  ---
@@ -167,5 +218,3 @@ tools_available:
167
218
 
168
219
  - If mode = `quick_prototype`: Load `./step-03-design.md`
169
220
  - If mode = `production`: Load `./step-02-scope.md`
170
-
171
- Remember: Do NOT proceed until user explicitly selects [C]
@@ -2,10 +2,10 @@
2
2
 
3
3
  ## MANDATORY EXECUTION RULES (READ FIRST)
4
4
 
5
- - 🛑 NEVER proceed without clear scope confirmation
6
- - 📖 CRITICAL: Pull relevant context from existing artifacts
7
- - ALWAYS confirm scope before design phase
8
- - 🎯 Goal: Define exactly what to design and gather visual references
5
+ - NEVER proceed without clear scope confirmation
6
+ - CRITICAL: Pull relevant context from existing artifacts
7
+ - ALWAYS confirm scope before design phase
8
+ - Goal: Define exactly what to design and gather visual references
9
9
 
10
10
  ---
11
11
 
@@ -16,6 +16,7 @@ You should have:
16
16
  - `user_intent`: What user wants to design
17
17
  - `project_state`: greenfield or existing
18
18
  - `related_artifacts`: List of relevant files
19
+ - `theme`: Theme info (if detected in Step 1)
19
20
 
20
21
  ---
21
22
 
@@ -126,7 +127,6 @@ Would you like to gather visual references before designing?
126
127
 
127
128
  **If U (user has URLs):**
128
129
  - Accept user's reference URLs
129
- - If Playwright available, offer to screenshot them
130
130
  - Note key elements to incorporate
131
131
 
132
132
  **If N:**
@@ -155,21 +155,6 @@ Ready to proceed to design phase.
155
155
 
156
156
  ---
157
157
 
158
- ## COLLABORATION MENU
159
-
160
- ```
161
- [A] Advanced - Deeper exploration of requirements
162
- [P] Party Mode - Get perspectives from PM, Dev, UX agents
163
- [C] Continue - Proceed to design phase
164
- ```
165
-
166
- **Menu Handlers:**
167
- - **A**: Explore requirements more deeply, ask clarifying questions
168
- - **P**: Load party mode for multi-agent input on design direction
169
- - **C**: Proceed to Step 3
170
-
171
- ---
172
-
173
158
  ## STATE TO CARRY FORWARD
174
159
 
175
160
  Add to working memory:
@@ -193,6 +178,4 @@ design_direction: "[summary of approach]"
193
178
 
194
179
  ## NEXT STEP
195
180
 
196
- When user selects [C], load `./step-03-design.md`
197
-
198
- Remember: Do NOT proceed until user explicitly selects [C]
181
+ Proceed to `./step-03-design.md`
@@ -2,10 +2,10 @@
2
2
 
3
3
  ## MANDATORY EXECUTION RULES (READ FIRST)
4
4
 
5
- - 🛑 NEVER proceed without user confirmation at each design iteration
6
- - 📖 CRITICAL: Load the appropriate tool file for detailed execution
7
- - ALWAYS show design results for review before moving on
8
- - 🎯 Goal: Create visual prototype using user's preferred tool
5
+ - NEVER proceed without user confirmation at each design iteration
6
+ - CRITICAL: Load the appropriate tool file for detailed execution
7
+ - ALWAYS show design results for review before moving on
8
+ - Goal: Create visual prototype using user's preferred tool
9
9
 
10
10
  ---
11
11
 
@@ -17,6 +17,7 @@ You should have:
17
17
  - `scope`: Specific items to design (if production mode)
18
18
  - `inspiration`: Visual references (if gathered)
19
19
  - `tools_available`: Which tools are configured
20
+ - `theme`: Theme info (if detected in Step 1)
20
21
 
21
22
  ---
22
23
 
@@ -38,25 +39,25 @@ DESIGN TOOL SELECTION
38
39
  Choose your design approach:
39
40
 
40
41
  [S] SuperDesign
41
- Quick HTML/CSS prototypes via VS Code
42
- Best for: Rapid visual exploration, custom styling
42
+ -> Quick HTML/CSS prototypes via VS Code
43
+ -> Best for: Rapid visual exploration, custom styling
43
44
 
44
45
  [T] Stitch (Google AI Design) {show availability status}
45
- AI-generated complete page designs
46
- Best for: Full page layouts, design system consistency
47
- Requires: Stitch MCP + Google's stitch-skills installed
46
+ -> AI-generated complete page designs
47
+ -> Best for: Full page layouts, design system consistency
48
+ -> Requires: Stitch MCP configured
48
49
 
49
50
  [M] MagicPatterns {show availability status}
50
- AI-generated React components
51
- Best for: Direct React code, component variations
51
+ -> AI-generated React components
52
+ -> Best for: Direct React code, component variations
52
53
 
53
54
  [W] Wireframe Only
54
- ASCII or Excalidraw sketch
55
- Best for: Structure-first, discussing layout
55
+ -> ASCII or Excalidraw sketch
56
+ -> Best for: Structure-first, discussing layout
56
57
 
57
58
  [D] Direct to Components
58
- Skip visuals, map directly to shadcn
59
- Best for: Standard patterns, known layouts
59
+ -> Skip visuals, map directly to shadcn
60
+ -> Best for: Standard patterns, known layouts
60
61
 
61
62
  Which approach? [S/T/M/W/D]
62
63
  ```
@@ -81,6 +82,7 @@ Pass context to tool execution:
81
82
  - `user_intent`
82
83
  - `scope` (if available)
83
84
  - `inspiration` (if gathered)
85
+ - `theme` (if detected in Step 1)
84
86
 
85
87
  ---
86
88
 
@@ -93,28 +95,7 @@ Execute the loaded tool file completely:
93
95
 
94
96
  ---
95
97
 
96
- ### 4. Offer Playwright Verification (Optional)
97
-
98
- If Playwright MCP is available AND user created a visual prototype (SuperDesign or MagicPatterns):
99
-
100
- ```
101
- VISUAL VERIFICATION (Optional)
102
-
103
- Would you like me to capture screenshots of your design?
104
-
105
- [Y] Yes - Take screenshots (desktop + mobile)
106
- [N] No - Skip verification
107
- ```
108
-
109
- If Y:
110
- - Use Playwright to navigate to design
111
- - Capture desktop screenshot
112
- - Resize viewport (375px width) and capture mobile screenshot
113
- - Present screenshots for review
114
-
115
- ---
116
-
117
- ### 5. Confirm Design Completion
98
+ ### 4. Confirm Design Completion
118
99
 
119
100
  After tool execution completes:
120
101
 
@@ -125,6 +106,12 @@ Tool used: {design.tool_used}
125
106
  Output: {design.output_location}
126
107
  ```
127
108
 
109
+ Direct user to review designs on the tool's platform:
110
+ - **MagicPatterns:** Review at the MagicPatterns URL
111
+ - **Stitch:** Review on the Stitch platform
112
+ - **SuperDesign:** Review in VS Code / browser
113
+ - **Wireframe/Direct:** Review inline
114
+
128
115
  **If Quick Prototype mode:**
129
116
  ```
130
117
  Prototype complete! Would you like to:
@@ -140,7 +127,7 @@ Ready to create dev handover artifacts.
140
127
 
141
128
  ---
142
129
 
143
- ### 5.5 Validate All Designs (Before Step 4)
130
+ ### 5. Validate All Designs (Before Step 4)
144
131
 
145
132
  **When scope has multiple items**, validate coverage before proceeding to artifacts:
146
133
 
@@ -149,10 +136,9 @@ DESIGN VALIDATION
149
136
 
150
137
  Checking scope coverage:
151
138
  {foreach scope_item}
152
- [✓/✗] {item_name}: {url_or_missing}
139
+ [done/missing] {item_name}: {url_or_missing}
153
140
  {/foreach}
154
141
 
155
- ─────────────────────────────────────
156
142
  Coverage: {completed_count} of {total_count} scope items
157
143
  ```
158
144
 
@@ -189,23 +175,10 @@ Options:
189
175
 
190
176
  **If all items covered:**
191
177
  ```
192
- VALIDATION PASSED
178
+ VALIDATION PASSED
193
179
 
194
180
  All {total_count} scope items have approved designs.
195
181
  Ready for artifact generation.
196
-
197
- [C] Continue to artifacts
198
- [R] Review designs before continuing
199
- ```
200
-
201
- ---
202
-
203
- ## COLLABORATION MENU
204
-
205
- ```
206
- [A] Advanced - Explore design alternatives
207
- [P] Party Mode - Get feedback from multiple perspectives
208
- [C] Continue - Proceed to next step
209
182
  ```
210
183
 
211
184
  ---
@@ -216,7 +189,7 @@ Should have from tool execution:
216
189
 
217
190
  ```yaml
218
191
  design:
219
- tool_used: [superdesign | magicpatterns | wireframe | direct]
192
+ tool_used: [superdesign | stitch | magicpatterns | wireframe | direct]
220
193
  output_location: "{path or URL}"
221
194
  output_format: [html | react | ascii | mapping]
222
195
  needs_conversion: [true | false]
@@ -229,6 +202,4 @@ design:
229
202
 
230
203
  - If `mode` = `quick_prototype` AND user selects [D]: End workflow
231
204
  - If `mode` = `quick_prototype` AND user selects [P]: Load `./step-04-artifacts.md`
232
- - If `mode` = `production` AND user selects [C]: Load `./step-04-artifacts.md`
233
-
234
- Remember: Do NOT proceed until user explicitly confirms.
205
+ - If `mode` = `production`: Load `./step-04-artifacts.md`