codingbuddy-rules 1.2.0 → 1.3.1
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.
- package/.ai-rules/adapters/opencode-skills.md +263 -0
- package/.ai-rules/adapters/opencode.md +517 -0
- package/.ai-rules/agents/README.md +114 -5
- package/.ai-rules/agents/act-mode.json +158 -0
- package/.ai-rules/agents/eval-mode.json +175 -0
- package/.ai-rules/agents/plan-mode.json +103 -0
- package/.ai-rules/keyword-modes.json +9 -3
- package/package.json +1 -1
|
@@ -0,0 +1,263 @@
|
|
|
1
|
+
# OpenCode Skills Integration Guide
|
|
2
|
+
|
|
3
|
+
This guide explains how to integrate `.ai-rules/skills/` with OpenCode/Crush's Agent Skills system.
|
|
4
|
+
|
|
5
|
+
## Overview
|
|
6
|
+
|
|
7
|
+
OpenCode/Crush supports Agent Skills standard for structured AI capabilities. This guide maps existing `.ai-rules/skills/` to OpenCode-compatible format and provides usage patterns.
|
|
8
|
+
|
|
9
|
+
## Skills Mapping
|
|
10
|
+
|
|
11
|
+
| .ai-rules Skill | OpenCode Usage | Agent Integration |
|
|
12
|
+
|----------------|----------------|-------------------|
|
|
13
|
+
| `brainstorming` | `/skill brainstorming` | plan agent |
|
|
14
|
+
| `test-driven-development` | `/skill tdd` | build agent |
|
|
15
|
+
| `systematic-debugging` | `/skill debug` | build agent |
|
|
16
|
+
| `writing-plans` | `/skill planning` | plan agent |
|
|
17
|
+
| `executing-plans` | `/skill execute` | build agent |
|
|
18
|
+
| `frontend-design` | `/skill frontend` | build agent |
|
|
19
|
+
| `dispatching-parallel-agents` | `/skill parallel` | plan agent |
|
|
20
|
+
| `subagent-driven-development` | `/skill subagent` | plan agent |
|
|
21
|
+
|
|
22
|
+
## Skills Configuration
|
|
23
|
+
|
|
24
|
+
### For OpenCode (.opencode.json)
|
|
25
|
+
```json
|
|
26
|
+
{
|
|
27
|
+
"skills": {
|
|
28
|
+
"paths": [
|
|
29
|
+
"packages/rules/.ai-rules/skills"
|
|
30
|
+
],
|
|
31
|
+
"auto_load": true
|
|
32
|
+
}
|
|
33
|
+
}
|
|
34
|
+
```
|
|
35
|
+
|
|
36
|
+
### For Crush (crush.json)
|
|
37
|
+
```json
|
|
38
|
+
{
|
|
39
|
+
"options": {
|
|
40
|
+
"skills_paths": [
|
|
41
|
+
"packages/rules/.ai-rules/skills"
|
|
42
|
+
],
|
|
43
|
+
"auto_suggest_skills": true
|
|
44
|
+
}
|
|
45
|
+
}
|
|
46
|
+
```
|
|
47
|
+
|
|
48
|
+
## Usage Patterns
|
|
49
|
+
|
|
50
|
+
### 1. Direct Skill Invocation
|
|
51
|
+
```bash
|
|
52
|
+
# In OpenCode CLI
|
|
53
|
+
/skill brainstorming "새로운 기능 아이디어"
|
|
54
|
+
/skill tdd "사용자 인증 구현"
|
|
55
|
+
/skill debug "로그인 버그 해결"
|
|
56
|
+
```
|
|
57
|
+
|
|
58
|
+
### 2. Agent + Skill Combination
|
|
59
|
+
```bash
|
|
60
|
+
# Planning with brainstorming skill
|
|
61
|
+
/agent plan
|
|
62
|
+
/skill brainstorming "대시보드 UI 개선"
|
|
63
|
+
|
|
64
|
+
# Implementation with TDD skill
|
|
65
|
+
/agent build
|
|
66
|
+
/skill tdd "API 연동 구현"
|
|
67
|
+
|
|
68
|
+
# Review with debugging skill
|
|
69
|
+
/agent reviewer
|
|
70
|
+
/skill debug "성능 이슈 분석"
|
|
71
|
+
```
|
|
72
|
+
|
|
73
|
+
### 3. Automatic Skill Recommendation
|
|
74
|
+
|
|
75
|
+
OpenCode can automatically recommend skills based on prompts using the `recommend_skills` MCP tool:
|
|
76
|
+
|
|
77
|
+
```typescript
|
|
78
|
+
// Auto-triggered when user enters certain keywords
|
|
79
|
+
"버그가 있어" → recommends: systematic-debugging
|
|
80
|
+
"계획을 세워줘" → recommends: writing-plans
|
|
81
|
+
"UI를 만들어줘" → recommends: frontend-design
|
|
82
|
+
```
|
|
83
|
+
|
|
84
|
+
## Skill Conversion Process
|
|
85
|
+
|
|
86
|
+
### 1. SKILL.md Format Compatibility
|
|
87
|
+
|
|
88
|
+
Existing `.ai-rules/skills/*/SKILL.md` files are already compatible with Agent Skills standard:
|
|
89
|
+
|
|
90
|
+
```markdown
|
|
91
|
+
# Skill Name
|
|
92
|
+
|
|
93
|
+
## Description
|
|
94
|
+
Brief description of the skill's purpose
|
|
95
|
+
|
|
96
|
+
## Usage
|
|
97
|
+
When and how to use this skill
|
|
98
|
+
|
|
99
|
+
## Steps
|
|
100
|
+
1. Step 1
|
|
101
|
+
2. Step 2
|
|
102
|
+
3. Step 3
|
|
103
|
+
|
|
104
|
+
## Examples
|
|
105
|
+
Example scenarios and outputs
|
|
106
|
+
```
|
|
107
|
+
|
|
108
|
+
### 2. No Conversion Required
|
|
109
|
+
|
|
110
|
+
The existing skills can be used directly in OpenCode without modification:
|
|
111
|
+
|
|
112
|
+
- **brainstorming/SKILL.md** ✅ Ready
|
|
113
|
+
- **test-driven-development/SKILL.md** ✅ Ready
|
|
114
|
+
- **systematic-debugging/SKILL.md** ✅ Ready
|
|
115
|
+
- **writing-plans/SKILL.md** ✅ Ready
|
|
116
|
+
- **executing-plans/SKILL.md** ✅ Ready
|
|
117
|
+
- **frontend-design/SKILL.md** ✅ Ready
|
|
118
|
+
- **dispatching-parallel-agents/SKILL.md** ✅ Ready
|
|
119
|
+
- **subagent-driven-development/SKILL.md** ✅ Ready
|
|
120
|
+
|
|
121
|
+
## Integration with MCP
|
|
122
|
+
|
|
123
|
+
### Codingbuddy MCP Skills
|
|
124
|
+
|
|
125
|
+
The `codingbuddy` MCP server provides skill recommendation:
|
|
126
|
+
|
|
127
|
+
```json
|
|
128
|
+
{
|
|
129
|
+
"mcp": {
|
|
130
|
+
"codingbuddy": {
|
|
131
|
+
"type": "stdio",
|
|
132
|
+
"command": ["npx", "codingbuddy@latest", "mcp"]
|
|
133
|
+
}
|
|
134
|
+
}
|
|
135
|
+
}
|
|
136
|
+
```
|
|
137
|
+
|
|
138
|
+
**Available MCP tools:**
|
|
139
|
+
- `recommend_skills`: Get skill recommendations based on user prompt
|
|
140
|
+
- `get_skill`: Load specific skill content
|
|
141
|
+
- `list_skills`: List all available skills
|
|
142
|
+
|
|
143
|
+
### Automatic Skill Loading
|
|
144
|
+
|
|
145
|
+
OpenCode agents can automatically load relevant skills:
|
|
146
|
+
|
|
147
|
+
```json
|
|
148
|
+
{
|
|
149
|
+
"agent": {
|
|
150
|
+
"plan": {
|
|
151
|
+
"auto_skills": ["brainstorming", "writing-plans"],
|
|
152
|
+
"systemPrompt": "You have access to brainstorming and planning skills..."
|
|
153
|
+
},
|
|
154
|
+
"build": {
|
|
155
|
+
"auto_skills": ["test-driven-development", "executing-plans", "frontend-design"],
|
|
156
|
+
"systemPrompt": "You have access to TDD, execution, and frontend skills..."
|
|
157
|
+
}
|
|
158
|
+
}
|
|
159
|
+
}
|
|
160
|
+
```
|
|
161
|
+
|
|
162
|
+
## Workflow Integration
|
|
163
|
+
|
|
164
|
+
### PLAN Mode Skills
|
|
165
|
+
When using `plan` agent, automatically suggest:
|
|
166
|
+
- **brainstorming**: For ideation and requirements exploration
|
|
167
|
+
- **writing-plans**: For structured implementation planning
|
|
168
|
+
- **dispatching-parallel-agents**: For complex multi-component features
|
|
169
|
+
|
|
170
|
+
### ACT Mode Skills
|
|
171
|
+
When using `build` agent, automatically suggest:
|
|
172
|
+
- **test-driven-development**: For core logic implementation
|
|
173
|
+
- **executing-plans**: For systematic plan execution
|
|
174
|
+
- **frontend-design**: For UI component development
|
|
175
|
+
- **systematic-debugging**: When encountering issues
|
|
176
|
+
|
|
177
|
+
### EVAL Mode Skills
|
|
178
|
+
When using `reviewer` agent, automatically suggest:
|
|
179
|
+
- **systematic-debugging**: For error analysis
|
|
180
|
+
- **subagent-driven-development**: For improvement strategies
|
|
181
|
+
|
|
182
|
+
## Korean Language Support
|
|
183
|
+
|
|
184
|
+
All skills support Korean language through MCP integration:
|
|
185
|
+
|
|
186
|
+
```bash
|
|
187
|
+
# Korean skill invocation
|
|
188
|
+
/skill 브레인스토밍 "새로운 기능"
|
|
189
|
+
/skill TDD개발 "사용자 관리"
|
|
190
|
+
/skill 디버깅 "성능 문제"
|
|
191
|
+
```
|
|
192
|
+
|
|
193
|
+
The MCP server handles Korean→English skill name mapping automatically.
|
|
194
|
+
|
|
195
|
+
## Examples
|
|
196
|
+
|
|
197
|
+
### Planning a New Feature
|
|
198
|
+
|
|
199
|
+
```bash
|
|
200
|
+
# Start planning session
|
|
201
|
+
/agent plan
|
|
202
|
+
|
|
203
|
+
# Use brainstorming skill
|
|
204
|
+
/skill brainstorming "사용자 대시보드 개선"
|
|
205
|
+
|
|
206
|
+
# Generate implementation plan
|
|
207
|
+
계획을 세워줘
|
|
208
|
+
```
|
|
209
|
+
|
|
210
|
+
### Implementing with TDD
|
|
211
|
+
|
|
212
|
+
```bash
|
|
213
|
+
# Switch to build agent
|
|
214
|
+
/agent build
|
|
215
|
+
|
|
216
|
+
# Load TDD skill
|
|
217
|
+
/skill test-driven-development "로그인 API 구현"
|
|
218
|
+
|
|
219
|
+
# Start TDD cycle
|
|
220
|
+
ACT
|
|
221
|
+
```
|
|
222
|
+
|
|
223
|
+
### Debugging Issues
|
|
224
|
+
|
|
225
|
+
```bash
|
|
226
|
+
# Use systematic debugging
|
|
227
|
+
/skill systematic-debugging "로그인 후 화면이 안 나와"
|
|
228
|
+
|
|
229
|
+
# Apply debugging methodology
|
|
230
|
+
디버깅해줘
|
|
231
|
+
```
|
|
232
|
+
|
|
233
|
+
## Benefits
|
|
234
|
+
|
|
235
|
+
### ✅ Advantages
|
|
236
|
+
- **No Migration Required**: Existing skills work as-is
|
|
237
|
+
- **Automatic Recommendations**: MCP-powered skill suggestions
|
|
238
|
+
- **Korean Support**: Full Korean language integration
|
|
239
|
+
- **Agent Integration**: Skills work seamlessly with agent modes
|
|
240
|
+
- **Workflow Enhancement**: Skills enhance PLAN/ACT/EVAL workflow
|
|
241
|
+
|
|
242
|
+
### ✅ Enhanced Capabilities
|
|
243
|
+
- **Context-Aware**: Skills adapt to current agent mode
|
|
244
|
+
- **Progressive Enhancement**: Skills complement agent capabilities
|
|
245
|
+
- **Structured Approach**: Consistent methodology across skills
|
|
246
|
+
- **Quality Focus**: Skills reinforce .ai-rules standards
|
|
247
|
+
|
|
248
|
+
## Maintenance
|
|
249
|
+
|
|
250
|
+
### Adding New Skills
|
|
251
|
+
|
|
252
|
+
1. Create new skill directory: `packages/rules/.ai-rules/skills/new-skill/`
|
|
253
|
+
2. Add `SKILL.md` file following Agent Skills format
|
|
254
|
+
3. Skills automatically available in OpenCode
|
|
255
|
+
4. Update skill mappings in this guide
|
|
256
|
+
|
|
257
|
+
### Updating Existing Skills
|
|
258
|
+
|
|
259
|
+
1. Modify skill content in `packages/rules/.ai-rules/skills/*/SKILL.md`
|
|
260
|
+
2. Changes automatically reflected in OpenCode
|
|
261
|
+
3. No configuration updates required
|
|
262
|
+
|
|
263
|
+
This integration provides seamless skills access within OpenCode while maintaining consistency with the broader `.ai-rules` ecosystem.
|