@neyugn/agent-kits 0.2.5 → 0.2.7

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/README.md CHANGED
@@ -163,7 +163,7 @@ The **Filter Skill** solves the "skill overload" problem by automatically detect
163
163
  | **1. Detection** | Scans for config files (`package.json`, `pubspec.yaml`, `Dockerfile`, etc.) |
164
164
  | **2. Recommendation** | Maps detected techstack to required skills |
165
165
  | **3. Confirmation** | Presents changes and asks about future techstack plans |
166
- | **4. Persistence** | Saves profile to `.agent/workspace-profile.json` |
166
+ | **4. Persistence** | Saves profile to `.agent/profile.json` |
167
167
 
168
168
  ### Example
169
169
 
package/README.vi.md CHANGED
@@ -155,7 +155,7 @@ Hoạt động trên **Windows**, **macOS**, và **Linux** với đường dẫn
155
155
  | **1. Phát hiện** | Quét các file config (`package.json`, `pubspec.yaml`, `Dockerfile`, etc.) |
156
156
  | **2. Đề xuất** | Map techstack đã phát hiện với các skills cần thiết |
157
157
  | **3. Xác nhận** | Hiển thị thay đổi và hỏi về kế hoạch techstack tương lai |
158
- | **4. Lưu trữ** | Lưu profile vào `.agent/workspace-profile.json` |
158
+ | **4. Lưu trữ** | Lưu profile vào `.agent/profile.json` |
159
159
 
160
160
  ### Ví dụ
161
161
 
package/README.zh.md CHANGED
@@ -155,7 +155,7 @@ npx @neyugn/agent-kits
155
155
  | **1. 检测** | 扫描配置文件(`package.json`、`pubspec.yaml`、`Dockerfile` 等) |
156
156
  | **2. 推荐** | 将检测到的技术栈映射到所需技能 |
157
157
  | **3. 确认** | 显示更改并询问未来的技术栈计划 |
158
- | **4. 持久化** | 将配置保存到 `.agent/workspace-profile.json` |
158
+ | **4. 持久化** | 将配置保存到 `.agent/profile.json` |
159
159
 
160
160
  ### 示例
161
161
 
@@ -0,0 +1,166 @@
1
+ # Common Skills Layer
2
+
3
+ > Universal skills shared across ALL kits in @neyugn/agent-kits
4
+
5
+ ---
6
+
7
+ ## 🎯 Purpose
8
+
9
+ The Common Skills Layer contains special skills that are shared across **all kits**. These skills:
10
+
11
+ 1. **Installed with every kit** - When a user installs any kit (coder, writer, etc.), common skills are installed automatically
12
+ 2. **Referenced in ARCHITECTURE.md** - Each kit's architecture document mentions these skills
13
+ 3. **Have dedicated workflows** - Invoked only when user calls the corresponding slash command
14
+
15
+ ---
16
+
17
+ ## 📁 Directory Structure
18
+
19
+ ```plaintext
20
+ common/
21
+ ├── COMMON.md # This file - documentation
22
+ ├── skills/ # Common skills
23
+ │ ├── scan-techstack/ # Techstack detection skill
24
+ │ │ ├── SKILL.md
25
+ │ │ └── scripts/
26
+ │ │ └── techstack_scanner.py
27
+ │ ├── filter-skill/ # Skill filtering skill
28
+ │ │ ├── SKILL.md
29
+ │ │ └── scripts/
30
+ │ │ └── workspace_analyzer.py
31
+ │ └── filter-agent/ # Agent filtering skill
32
+ │ └── SKILL.md
33
+ └── workflows/ # Common workflows
34
+ └── filter.md # /filter command
35
+ ```
36
+
37
+ ---
38
+
39
+ ## 🧩 Common Skills
40
+
41
+ | Skill | Description | Trigger |
42
+ | ---------------- | ------------------------------------------------------- | ---------------------- |
43
+ | `scan-techstack` | Analyze workspace to detect technologies and frameworks | Part of `/filter` flow |
44
+ | `filter-skill` | Recommend enable/disable skills based on techstack | Part of `/filter` flow |
45
+ | `filter-agent` | Recommend disable agents based on techstack | Part of `/filter` flow |
46
+
47
+ ---
48
+
49
+ ## 🔄 How It Works
50
+
51
+ ### Installation
52
+
53
+ When user runs `npx @neyugn/agent-kits`:
54
+
55
+ 1. User selects a kit (e.g., `coder`)
56
+ 2. Installer copies kit to workspace
57
+ 3. Installer copies `common/` skills to the same location
58
+ 4. Common skills are merged into the kit's architecture
59
+
60
+ ### Usage
61
+
62
+ ```bash
63
+ # User invokes workflow to filter skills and agents
64
+ /filter
65
+
66
+ # The workflow will:
67
+ # 1. scan-techstack: Analyze workspace (package.json, pubspec.yaml, etc.)
68
+ # 2. filter-skill: Recommend skill enable/disable based on techstack
69
+ # 3. filter-agent: Recommend agent disable based on techstack
70
+ # 4. Ask user confirmation + future techstack plans
71
+ # 5. Save results to .agent/profile.json
72
+ ```
73
+
74
+ ### Workflow Flow
75
+
76
+ ```
77
+ /filter
78
+
79
+ ├── Step 1: scan-techstack
80
+ │ └── Output: TechstackProfile (languages, frameworks, categories)
81
+
82
+ ├── Step 2: filter-skill
83
+ │ └── Output: SkillRecommendations (enable/disable lists)
84
+
85
+ ├── Step 3: filter-agent
86
+ │ └── Output: AgentRecommendations (disable list)
87
+
88
+ ├── Step 4: User Confirmation
89
+ │ └── Ask about future techstack plans
90
+
91
+ └── Step 5: Save to profile.json
92
+ ```
93
+
94
+ ---
95
+
96
+ ## 📊 Integration with Kits
97
+
98
+ Each kit's `ARCHITECTURE.md` MUST mention:
99
+
100
+ ```markdown
101
+ ## 🔗 Common Skills
102
+
103
+ This kit inherits from the **Common Skills Layer**. See `common/COMMON.md` for:
104
+
105
+ - `/filter` - Workspace-aware skill and agent filtering
106
+ - `scan-techstack` - Techstack detection
107
+ - `filter-skill` - Skill recommendations
108
+ - `filter-agent` - Agent recommendations
109
+
110
+ Common skills are automatically installed and available in all kits.
111
+ ```
112
+
113
+ ---
114
+
115
+ ## 📄 Profile Format
116
+
117
+ The `/filter` workflow saves results to `.agent/profile.json`:
118
+
119
+ ```json
120
+ {
121
+ "version": "1.0",
122
+ "generatedAt": "2026-02-05T12:00:00Z",
123
+ "analyzedBy": "filter-workflow v1.0",
124
+ "techstack": {
125
+ "languages": ["typescript", "python"],
126
+ "frameworks": ["nextjs", "tailwindcss"],
127
+ "databases": ["postgresql"],
128
+ "tools": ["docker", "github-actions"]
129
+ },
130
+ "skills": {
131
+ "enabled": ["react-patterns", "tailwind-patterns", "postgres-patterns"],
132
+ "disabled": ["flutter-patterns", "mobile-design", "queue-patterns"],
133
+ "userOverrides": {
134
+ "force-enabled": ["ai-rag-patterns"],
135
+ "force-disabled": []
136
+ }
137
+ },
138
+ "agents": {
139
+ "disabled": ["mobile-developer", "queue-specialist", "realtime-specialist"]
140
+ },
141
+ "futureTechstack": ["react-native", "kubernetes"]
142
+ }
143
+ ```
144
+
145
+ ---
146
+
147
+ ## 🚀 Future Common Skills (Planned)
148
+
149
+ | Skill | Description | Status |
150
+ | ----------------- | --------------------------------------------- | ------- |
151
+ | `context-manager` | Manage context length, auto-summarize history | Planned |
152
+ | `memory-skill` | Store and recall important information | Planned |
153
+ | `preference-sync` | Sync user preferences between sessions | Planned |
154
+
155
+ ---
156
+
157
+ ## 📝 Adding Common Skills
158
+
159
+ 1. Create skill folder in `common/skills/`
160
+ 2. Create SKILL.md with proper frontmatter
161
+ 3. (Optional) Create scripts/ for automation
162
+ 4. Update workflow in `common/workflows/` if needed
163
+ 5. Update this file's Skills table
164
+ 6. Update all kits' ARCHITECTURE.md to reference
165
+
166
+ ---
@@ -0,0 +1,194 @@
1
+ ---
2
+ name: filter-agent
3
+ description: Recommend enabling/disabling agents based on detected techstack. Reduces noise from irrelevant agents in the workspace.
4
+ category: common
5
+ trigger: manual
6
+ ---
7
+
8
+ # Filter Agent Skill
9
+
10
+ > Workspace-aware agent filtering for optimal AI performance.
11
+
12
+ ---
13
+
14
+ ## 🎯 Purpose
15
+
16
+ Filter Agent analyzes the **techstack profile** (from scan-techstack) and recommends which agents to **disable** based on project needs. This:
17
+
18
+ 1. **Reduces cognitive load** - AI focuses on relevant agents only
19
+ 2. **Improves accuracy** - No confusion from unrelated agent personas
20
+ 3. **Optimizes routing** - Orchestrator routes to appropriate specialists
21
+
22
+ ---
23
+
24
+ ## 🤖 Agent Categories
25
+
26
+ ### Tier 1: Master Agents (NEVER DISABLE)
27
+
28
+ | Agent | Always Required For |
29
+ | ----------------- | -------------------------------- |
30
+ | `orchestrator` | Multi-agent coordination |
31
+ | `project-planner` | Project planning, task breakdown |
32
+ | `debugger` | Systematic problem solving |
33
+
34
+ ### Tier 2: Development Specialists
35
+
36
+ | Agent | Enable When |
37
+ | --------------------- | ------------------------------------------ |
38
+ | `frontend-specialist` | Frontend detected (React, Vue, Angular) |
39
+ | `backend-specialist` | Backend detected (Express, NestJS, API) |
40
+ | `mobile-developer` | Mobile detected (Flutter, RN, iOS/Android) |
41
+ | `database-specialist` | Database detected (Prisma, PostgreSQL) |
42
+ | `devops-engineer` | DevOps detected (Docker, K8s, CI/CD) |
43
+
44
+ ### Tier 3: Quality & Security
45
+
46
+ | Agent | Enable When |
47
+ | --------------------- | ------------------------------------------- |
48
+ | `security-auditor` | Always enabled for security reviews |
49
+ | `code-reviewer` | Always enabled for PR reviews |
50
+ | `test-engineer` | Testing detected (Jest, Vitest, Playwright) |
51
+ | `performance-analyst` | Always enabled for optimization |
52
+
53
+ ### Tier 4: Domain Specialists
54
+
55
+ | Agent | Enable When |
56
+ | ------------------------ | ------------------------------------------ |
57
+ | `realtime-specialist` | Realtime detected (Socket.IO, WS) |
58
+ | `multi-tenant-architect` | Multi-tenancy patterns detected |
59
+ | `queue-specialist` | Queue detected (BullMQ, RabbitMQ) |
60
+ | `integration-specialist` | External API integrations detected |
61
+ | `ai-engineer` | AI detected (OpenAI, LangChain) |
62
+ | `cloud-architect` | Cloud infra detected (AWS, GCP, Terraform) |
63
+ | `data-engineer` | Data pipeline patterns detected |
64
+
65
+ ### Tier 5: Support Agents
66
+
67
+ | Agent | Enable When |
68
+ | ---------------------- | -------------------------- |
69
+ | `documentation-writer` | Always enabled for docs |
70
+ | `i18n-specialist` | i18n dependencies detected |
71
+ | `ux-researcher` | Frontend/Mobile detected |
72
+
73
+ ---
74
+
75
+ ## 📋 Mapping Rules
76
+
77
+ ### Category → Agent Mapping
78
+
79
+ | Detected Category | Agents to ENABLE |
80
+ | ----------------- | ------------------------------------------------ |
81
+ | `frontend` | frontend-specialist, ux-researcher |
82
+ | `backend` | backend-specialist |
83
+ | `mobile` | mobile-developer, ux-researcher |
84
+ | `database` | database-specialist |
85
+ | `devops` | devops-engineer, cloud-architect |
86
+ | `ai` | ai-engineer |
87
+ | `realtime` | realtime-specialist |
88
+ | `queue` | queue-specialist |
89
+ | `graphql` | backend-specialist (GraphQL expertise) |
90
+ | `auth` | security-auditor (always on), backend-specialist |
91
+ | `testing` | test-engineer |
92
+
93
+ ### Never Disable (Core Agents)
94
+
95
+ These agents are ALWAYS enabled regardless of techstack:
96
+
97
+ - `orchestrator` - Master coordinator
98
+ - `project-planner` - Planning and task management
99
+ - `debugger` - Problem solving
100
+ - `security-auditor` - Security reviews
101
+ - `code-reviewer` - Code quality
102
+ - `documentation-writer` - Documentation
103
+
104
+ ---
105
+
106
+ ## 📊 Recommendation Output
107
+
108
+ ```json
109
+ {
110
+ "agents": {
111
+ "enabled": [
112
+ "orchestrator",
113
+ "project-planner",
114
+ "debugger",
115
+ "frontend-specialist",
116
+ "backend-specialist",
117
+ "database-specialist",
118
+ "devops-engineer",
119
+ "test-engineer",
120
+ "security-auditor",
121
+ "code-reviewer",
122
+ "documentation-writer"
123
+ ],
124
+ "disabled": [
125
+ "mobile-developer",
126
+ "realtime-specialist",
127
+ "queue-specialist",
128
+ "ai-engineer",
129
+ "multi-tenant-architect",
130
+ "i18n-specialist",
131
+ "data-engineer"
132
+ ],
133
+ "coreAgents": [
134
+ "orchestrator",
135
+ "project-planner",
136
+ "debugger",
137
+ "security-auditor",
138
+ "code-reviewer",
139
+ "documentation-writer"
140
+ ]
141
+ }
142
+ }
143
+ ```
144
+
145
+ ---
146
+
147
+ ## 🔗 Integration
148
+
149
+ ### Input: TechstackProfile
150
+
151
+ Receives output from `scan-techstack`:
152
+
153
+ ```json
154
+ {
155
+ "categories": {
156
+ "frontend": true,
157
+ "backend": true,
158
+ "mobile": false,
159
+ "database": true,
160
+ "devops": true,
161
+ "ai": false,
162
+ "realtime": false,
163
+ "queue": false
164
+ }
165
+ }
166
+ ```
167
+
168
+ ### Output: AgentProfile
169
+
170
+ Produces agent recommendations for `profile.json`:
171
+
172
+ ```json
173
+ {
174
+ "agents": {
175
+ "disabled": [
176
+ "mobile-developer",
177
+ "realtime-specialist",
178
+ "queue-specialist",
179
+ "ai-engineer"
180
+ ]
181
+ }
182
+ }
183
+ ```
184
+
185
+ ---
186
+
187
+ ## ⚠️ Important Notes
188
+
189
+ 1. **Conservative approach** - When in doubt, keep agent enabled
190
+ 2. **User override** - User can force enable/disable any agent
191
+ 3. **Future planning** - If user plans to add mobile, keep mobile-developer enabled
192
+ 4. **Orchestrator respects profile** - Will skip disabled agents when routing
193
+
194
+ ---
@@ -0,0 +1,225 @@
1
+ ---
2
+ name: filter-skill
3
+ description: Recommend enabling/disabling skills based on detected techstack. Reduces noise from irrelevant skills in the workspace.
4
+ category: common
5
+ trigger: manual
6
+ workflow: /filter
7
+ ---
8
+
9
+ # Filter Skill
10
+
11
+ > Workspace-aware skill filtering for optimal AI performance.
12
+
13
+ ---
14
+
15
+ ## 🎯 Purpose
16
+
17
+ Filter Skill analyzes the **techstack profile** (from scan-techstack) and recommends which skills to **enable/disable** based on project needs. This:
18
+
19
+ 1. **Reduces noise** - Only relevant skills are loaded
20
+ 2. **Improves context** - AI focuses on applicable patterns
21
+ 3. **Optimizes performance** - Less skill content to process
22
+
23
+ ---
24
+
25
+ ## 🔗 Dependency
26
+
27
+ This skill requires **scan-techstack** to run first:
28
+
29
+ ```
30
+ /filter workflow:
31
+ 1. scan-techstack → detect techstack
32
+ 2. filter-skill → recommend skills (this skill)
33
+ 3. filter-agent → recommend agents
34
+ ```
35
+
36
+ ---
37
+
38
+ ## 📋 Skill Categories
39
+
40
+ ### Core Skills (NEVER DISABLE)
41
+
42
+ These skills are ALWAYS enabled regardless of techstack:
43
+
44
+ | Skill | Reason |
45
+ | ----------------------- | ----------------------------- |
46
+ | `clean-code` | Universal coding standards |
47
+ | `brainstorming` | Socratic questioning protocol |
48
+ | `plan-writing` | Task breakdown and WBS |
49
+ | `systematic-debugging` | 4-phase debugging methodology |
50
+ | `testing-patterns` | Testing pyramid, AAA pattern |
51
+ | `security-fundamentals` | OWASP 2025 security basics |
52
+
53
+ ### Frontend Skills
54
+
55
+ | Skill | Enable When |
56
+ | ------------------------ | ------------------------------- |
57
+ | `react-patterns` | React/Next.js detected |
58
+ | `typescript-patterns` | TypeScript detected |
59
+ | `tailwind-patterns` | Tailwind CSS detected |
60
+ | `frontend-design` | Any frontend framework |
61
+ | `seo-patterns` | Next.js, Nuxt.js (SSR) detected |
62
+ | `accessibility-patterns` | Frontend detected |
63
+
64
+ ### Backend Skills
65
+
66
+ | Skill | Enable When |
67
+ | ----------------------- | -------------------------- |
68
+ | `api-patterns` | Backend framework detected |
69
+ | `auth-patterns` | Auth dependencies detected |
70
+ | `graphql-patterns` | GraphQL detected |
71
+ | `nodejs-best-practices` | Node.js detected |
72
+
73
+ ### Database Skills
74
+
75
+ | Skill | Enable When |
76
+ | ------------------- | --------------------- |
77
+ | `database-design` | Any database detected |
78
+ | `postgres-patterns` | PostgreSQL detected |
79
+ | `redis-patterns` | Redis detected |
80
+
81
+ ### Mobile Skills
82
+
83
+ | Skill | Enable When |
84
+ | ----------------------- | --------------------- |
85
+ | `flutter-patterns` | Flutter/Dart detected |
86
+ | `react-native-patterns` | React Native detected |
87
+ | `mobile-design` | Any mobile platform |
88
+
89
+ ### DevOps Skills
90
+
91
+ | Skill | Enable When |
92
+ | -------------------------- | -------------------------- |
93
+ | `docker-patterns` | Docker detected |
94
+ | `kubernetes-patterns` | Kubernetes detected |
95
+ | `terraform-patterns` | Terraform detected |
96
+ | `github-actions` | GitHub Actions detected |
97
+ | `gitlab-ci-patterns` | GitLab CI detected |
98
+ | `monitoring-observability` | DevOps/Production detected |
99
+
100
+ ### AI Skills
101
+
102
+ | Skill | Enable When |
103
+ | -------------------- | ---------------------------- |
104
+ | `ai-rag-patterns` | AI/LLM dependencies detected |
105
+ | `prompt-engineering` | AI/LLM dependencies detected |
106
+
107
+ ### Realtime & Queue Skills
108
+
109
+ | Skill | Enable When |
110
+ | ------------------- | ------------------------------ |
111
+ | `realtime-patterns` | Socket.IO/WebSocket detected |
112
+ | `queue-patterns` | BullMQ/RabbitMQ detected |
113
+ | `multi-tenancy` | Multi-tenant patterns detected |
114
+
115
+ ### Support Skills
116
+
117
+ | Skill | Enable When |
118
+ | ------------------------- | -------------------------- |
119
+ | `i18n-localization` | i18n dependencies detected |
120
+ | `documentation-templates` | Always available |
121
+ | `mermaid-diagrams` | Always available |
122
+
123
+ ---
124
+
125
+ ## 📊 Recommendation Output
126
+
127
+ ```json
128
+ {
129
+ "skills": {
130
+ "enabled": [
131
+ "clean-code",
132
+ "testing-patterns",
133
+ "security-fundamentals",
134
+ "react-patterns",
135
+ "typescript-patterns",
136
+ "tailwind-patterns",
137
+ "frontend-design",
138
+ "api-patterns",
139
+ "database-design",
140
+ "postgres-patterns",
141
+ "docker-patterns",
142
+ "github-actions"
143
+ ],
144
+ "disabled": [
145
+ "flutter-patterns",
146
+ "react-native-patterns",
147
+ "mobile-design",
148
+ "queue-patterns",
149
+ "realtime-patterns",
150
+ "ai-rag-patterns",
151
+ "kubernetes-patterns"
152
+ ],
153
+ "coreSkills": [
154
+ "clean-code",
155
+ "brainstorming",
156
+ "plan-writing",
157
+ "systematic-debugging",
158
+ "testing-patterns",
159
+ "security-fundamentals"
160
+ ]
161
+ }
162
+ }
163
+ ```
164
+
165
+ ---
166
+
167
+ ## 🔧 Mapping Rules
168
+
169
+ ### Category → Skill Mapping
170
+
171
+ | Detected Category | Skills to ENABLE |
172
+ | ----------------- | ------------------------------------------------------------------------------------------------------------- |
173
+ | `frontend` | react-patterns, typescript-patterns, frontend-design, tailwind-patterns, seo-patterns, accessibility-patterns |
174
+ | `backend` | api-patterns, nodejs-best-practices, auth-patterns |
175
+ | `mobile` | flutter-patterns OR react-native-patterns, mobile-design |
176
+ | `database` | database-design, postgres-patterns OR redis-patterns |
177
+ | `devops` | docker-patterns, kubernetes-patterns, github-actions, monitoring-observability |
178
+ | `ai` | ai-rag-patterns, prompt-engineering |
179
+ | `realtime` | realtime-patterns |
180
+ | `queue` | queue-patterns |
181
+ | `graphql` | graphql-patterns |
182
+
183
+ ### Framework → Additional Skills
184
+
185
+ | Framework | Additional Skills |
186
+ | ------------- | ---------------------------------- |
187
+ | `nextjs` | seo-patterns, react-patterns |
188
+ | `tailwindcss` | tailwind-patterns |
189
+ | `prisma` | database-design, postgres-patterns |
190
+ | `socketio` | realtime-patterns |
191
+ | `gitlab-ci` | gitlab-ci-patterns |
192
+ | `terraform` | terraform-patterns |
193
+
194
+ ---
195
+
196
+ ## ⚠️ Important Notes
197
+
198
+ 1. **Core skills always ON** - Never disable core skills
199
+ 2. **Conservative approach** - When in doubt, keep skill enabled
200
+ 3. **User override** - User can force enable/disable any skill
201
+ 4. **Future planning** - Check user's future techstack plans before disabling
202
+
203
+ ---
204
+
205
+ ## 📄 Persistence
206
+
207
+ Results are saved to `.agent/profile.json`:
208
+
209
+ ```json
210
+ {
211
+ "version": "1.0",
212
+ "generatedAt": "2026-02-05T12:00:00Z",
213
+ "analyzedBy": "filter-skill v1.0",
214
+ "skills": {
215
+ "enabled": [...],
216
+ "disabled": [...],
217
+ "userOverrides": {
218
+ "force-enabled": [],
219
+ "force-disabled": []
220
+ }
221
+ }
222
+ }
223
+ ```
224
+
225
+ ---