start-vibing 3.0.10 → 4.0.0

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.
@@ -1,208 +1,208 @@
1
- # Configuracao de Agentes - AI Builder
2
-
3
- Este diretorio contem a configuracao completa do sistema de agentes
4
- para o fluxo de desenvolvimento do AI Builder.
5
-
6
- ---
7
-
8
- ## Estrutura
9
-
10
- ```
11
- .claude/
12
- ├── CLAUDE.md # Contexto geral para agentes
13
- ├── settings.json # Configuracoes globais
14
- ├── README.md # Este arquivo
15
-
16
- ├── agents/ # Subagentes especializados
17
- │ ├── orchestrator.md # Coordena o fluxo
18
- │ ├── analyzer.md # Analisa impacto
19
- │ ├── documenter.md # Cria documentacao
20
- │ ├── tester.md # Cria e executa testes
21
- │ ├── security-auditor.md # Audita seguranca
22
- │ ├── ui-ux-reviewer.md # Revisa UI/UX
23
- │ ├── quality-checker.md # Verifica qualidade
24
- │ └── final-validator.md # Validacao final
25
-
26
- ├── hooks/ # Hooks de seguranca
27
- │ └── security-check.js # Bloqueia acoes perigosas
28
-
29
- └── commands/ # Slash commands customizados
30
- ├── feature.md # /feature - nova feature
31
- ├── fix.md # /fix - corrigir bug
32
- ├── validate.md # /validate - validacao completa
33
- └── research.md # /research - pesquisar topico
34
- ```
35
-
36
- ---
37
-
38
- ## Agentes
39
-
40
- ### Hierarquia
41
-
42
- ```
43
- ┌─────────────────┐
44
- │ ORCHESTRATOR │
45
- │ (coordenador) │
46
- └────────┬────────┘
47
-
48
- ┌────────────────────┼────────────────────┐
49
- │ │ │
50
- ┌────▼────┐ ┌────▼────┐ ┌────▼────┐
51
- │ANALYZER │ │ TESTER │ │ UI/UX │
52
- │(impacto)│ │(testes) │ │REVIEWER │
53
- └─────────┘ └─────────┘ └─────────┘
54
- │ │ │
55
- ┌────▼────┐ ┌────▼────┐ ┌────▼────┐
56
- │DOCUMENT │ │SECURITY │ │ QUALITY │
57
- │ ER │ │ AUDITOR │ │ CHECKER │
58
- └─────────┘ └─────────┘ └─────────┘
59
-
60
- ┌────────▼────────┐
61
- │ FINAL VALIDATOR │
62
- │ (veto power) │
63
- └─────────────────┘
64
- ```
65
-
66
- ### Funcoes
67
-
68
- | Agente | Funcao | Pesquisa? | Pode Vetar? |
69
- | ---------------- | -------------------------------- | --------------------------------------- | ----------- |
70
- | orchestrator | Coordena fluxo, delega tarefas | **SIM** - best practices gerais | Nao |
71
- | analyzer | Analisa impacto de mudancas | **SIM** - arquitetura, migracoes | Nao |
72
- | documenter | Cria/atualiza documentacao | **SIM** - padroes de docs | Nao |
73
- | tester | Cria e executa testes | **SIM** - padroes de teste, edge cases | Nao |
74
- | security-auditor | Valida seguranca | **SIM** - OWASP, CVEs, vulnerabilidades | **SIM** |
75
- | ui-ux-reviewer | Pesquisa competidores, valida UI | **SIM** - competidores, UX, WCAG | Nao |
76
- | quality-checker | Roda typecheck/lint/build | **SIM** - solucoes para erros | Nao |
77
- | final-validator | Checklist final completo | **SIM** - valida que outros pesquisaram | **SIM** |
78
-
79
- ### TODOS os Agentes Pesquisam
80
-
81
- Cada agente tem a obrigacao de pesquisar best practices antes de agir:
82
-
83
- ```
84
- orchestrator → Pesquisa padroes de arquitetura e implementacao
85
- analyzer → Pesquisa riscos e migracoes
86
- documenter → Pesquisa padroes de documentacao
87
- tester → Pesquisa edge cases e padroes de teste
88
- security-auditor → Pesquisa OWASP, CVEs, vulnerabilidades conhecidas
89
- ui-ux-reviewer → Pesquisa competidores, UX patterns, WCAG
90
- quality-checker → Pesquisa solucoes para erros encontrados
91
- final-validator → Valida que TODOS os outros pesquisaram
92
- ```
93
-
94
- ---
95
-
96
- ## Fluxos
97
-
98
- ### Feature Completa (default_flow)
99
-
100
- ```
101
- 1. analyzer → Analisa impacto
102
- 2. ui-ux-reviewer → Pesquisa competidores (se UI)
103
- 3. documenter → Documenta fluxo
104
- 4. [IMPLEMENTACAO]
105
- 5. tester → Cria testes
106
- 6. security-auditor → Valida seguranca
107
- 7. quality-checker → typecheck + lint + build
108
- 8. final-validator → Checklist final
109
- 9. [COMMIT]
110
- ```
111
-
112
- ### Bug Fix (bug_fix_flow)
113
-
114
- ```
115
- 1. analyzer → Identifica causa raiz
116
- 2. [IMPLEMENTACAO]
117
- 3. tester → Cria teste de regressao
118
- 4. security-auditor → Valida que fix nao introduz vuln
119
- 5. quality-checker → typecheck + lint + build
120
- 6. final-validator → Checklist adaptado
121
- 7. [COMMIT]
122
- ```
123
-
124
- ### Refactor (refactor_flow)
125
-
126
- ```
127
- 1. analyzer → Mapeia dependencias
128
- 2. [IMPLEMENTACAO]
129
- 3. tester → Atualiza testes
130
- 4. security-auditor → Valida seguranca
131
- 5. quality-checker → typecheck + lint + build
132
- 6. final-validator → Checklist adaptado
133
- 7. [COMMIT]
134
- ```
135
-
136
- ### Config/Chore (config_flow)
137
-
138
- ```
139
- 1. quality-checker → Apenas qualidade
140
- 2. [COMMIT]
141
- ```
142
-
143
- ---
144
-
145
- ## Comandos
146
-
147
- | Comando | Descricao | Uso |
148
- | ----------- | ------------------- | ------------------------- |
149
- | `/feature` | Inicia nova feature | `/feature [descricao]` |
150
- | `/fix` | Corrige bug | `/fix [descricao do bug]` |
151
- | `/validate` | Validacao completa | `/validate` |
152
- | `/research` | Pesquisa topico | `/research [topico]` |
153
-
154
- ---
155
-
156
- ## Hooks
157
-
158
- ### security-check.js
159
-
160
- Executado ANTES de cada tool call. Bloqueia:
161
-
162
- - Comandos destrutivos (`rm -rf`, `sudo`, etc)
163
- - Codigo inseguro (`eval()`, `innerHTML`, etc)
164
- - Dados sensiveis hardcoded (passwords, API keys)
165
- - Violacoes de regras do projeto (userId do request, z.any(), etc)
166
-
167
- ---
168
-
169
- ## Como Usar
170
-
171
- ### No Claude Code (Interativo)
172
-
173
- Os agentes sao carregados automaticamente quando voce abre o projeto.
174
- O orchestrator coordena o fluxo baseado no tipo de tarefa.
175
-
176
- ### Com Claude Agent SDK (Programatico)
177
-
178
- ```python
179
- from claude_agent_sdk import ClaudeSDKClient
180
-
181
- client = ClaudeSDKClient(
182
- setting_sources=["project"] # Carrega .claude/settings.json
183
- )
184
-
185
- result = await client.query("Implementar feature X")
186
- ```
187
-
188
- ---
189
-
190
- ## Regras Garantidas pelos Agentes
191
-
192
- 1. **Seguranca** - security-auditor valida TODAS as regras
193
- 2. **Testes** - tester garante cobertura adequada
194
- 3. **Documentacao** - documenter cria docs obrigatorias
195
- 4. **UI/UX** - ui-ux-reviewer pesquisa competidores
196
- 5. **Qualidade** - quality-checker roda verificacoes
197
- 6. **Conformidade** - final-validator valida TUDO antes do commit
198
-
199
- ---
200
-
201
- ## Atualizacao
202
-
203
- Ao adicionar novas regras ao projeto:
204
-
205
- 1. Atualizar `/claude.md` com a regra
206
- 2. Atualizar agente relevante em `.claude/agents/`
207
- 3. Atualizar `final-validator.md` com novo item no checklist
208
- 4. Se necessario, adicionar pattern em `hooks/security-check.js`
1
+ # Configuracao de Agentes - AI Builder
2
+
3
+ Este diretorio contem a configuracao completa do sistema de agentes
4
+ para o fluxo de desenvolvimento do AI Builder.
5
+
6
+ ---
7
+
8
+ ## Estrutura
9
+
10
+ ```
11
+ .claude/
12
+ ├── CLAUDE.md # Contexto geral para agentes
13
+ ├── settings.json # Configuracoes globais
14
+ ├── README.md # Este arquivo
15
+
16
+ ├── agents/ # Subagentes especializados
17
+ │ ├── orchestrator.md # Coordena o fluxo
18
+ │ ├── analyzer.md # Analisa impacto
19
+ │ ├── documenter.md # Cria documentacao
20
+ │ ├── tester.md # Cria e executa testes
21
+ │ ├── security-auditor.md # Audita seguranca
22
+ │ ├── ui-ux-reviewer.md # Revisa UI/UX
23
+ │ ├── quality-checker.md # Verifica qualidade
24
+ │ └── final-validator.md # Validacao final
25
+
26
+ ├── hooks/ # Hooks de seguranca
27
+ │ └── security-check.js # Bloqueia acoes perigosas
28
+
29
+ └── commands/ # Slash commands customizados
30
+ ├── feature.md # /feature - nova feature
31
+ ├── fix.md # /fix - corrigir bug
32
+ ├── validate.md # /validate - validacao completa
33
+ └── research.md # /research - pesquisar topico
34
+ ```
35
+
36
+ ---
37
+
38
+ ## Agentes
39
+
40
+ ### Hierarquia
41
+
42
+ ```
43
+ ┌─────────────────┐
44
+ │ ORCHESTRATOR │
45
+ │ (coordenador) │
46
+ └────────┬────────┘
47
+
48
+ ┌────────────────────┼────────────────────┐
49
+ │ │ │
50
+ ┌────▼────┐ ┌────▼────┐ ┌────▼────┐
51
+ │ANALYZER │ │ TESTER │ │ UI/UX │
52
+ │(impacto)│ │(testes) │ │REVIEWER │
53
+ └─────────┘ └─────────┘ └─────────┘
54
+ │ │ │
55
+ ┌────▼────┐ ┌────▼────┐ ┌────▼────┐
56
+ │DOCUMENT │ │SECURITY │ │ QUALITY │
57
+ │ ER │ │ AUDITOR │ │ CHECKER │
58
+ └─────────┘ └─────────┘ └─────────┘
59
+
60
+ ┌────────▼────────┐
61
+ │ FINAL VALIDATOR │
62
+ │ (veto power) │
63
+ └─────────────────┘
64
+ ```
65
+
66
+ ### Funcoes
67
+
68
+ | Agente | Funcao | Pesquisa? | Pode Vetar? |
69
+ | ---------------- | -------------------------------- | --------------------------------------- | ----------- |
70
+ | orchestrator | Coordena fluxo, delega tarefas | **SIM** - best practices gerais | Nao |
71
+ | analyzer | Analisa impacto de mudancas | **SIM** - arquitetura, migracoes | Nao |
72
+ | documenter | Cria/atualiza documentacao | **SIM** - padroes de docs | Nao |
73
+ | tester | Cria e executa testes | **SIM** - padroes de teste, edge cases | Nao |
74
+ | security-auditor | Valida seguranca | **SIM** - OWASP, CVEs, vulnerabilidades | **SIM** |
75
+ | ui-ux-reviewer | Pesquisa competidores, valida UI | **SIM** - competidores, UX, WCAG | Nao |
76
+ | quality-checker | Roda typecheck/lint/build | **SIM** - solucoes para erros | Nao |
77
+ | final-validator | Checklist final completo | **SIM** - valida que outros pesquisaram | **SIM** |
78
+
79
+ ### TODOS os Agentes Pesquisam
80
+
81
+ Cada agente tem a obrigacao de pesquisar best practices antes de agir:
82
+
83
+ ```
84
+ orchestrator → Pesquisa padroes de arquitetura e implementacao
85
+ analyzer → Pesquisa riscos e migracoes
86
+ documenter → Pesquisa padroes de documentacao
87
+ tester → Pesquisa edge cases e padroes de teste
88
+ security-auditor → Pesquisa OWASP, CVEs, vulnerabilidades conhecidas
89
+ ui-ux-reviewer → Pesquisa competidores, UX patterns, WCAG
90
+ quality-checker → Pesquisa solucoes para erros encontrados
91
+ final-validator → Valida que TODOS os outros pesquisaram
92
+ ```
93
+
94
+ ---
95
+
96
+ ## Fluxos
97
+
98
+ ### Feature Completa (default_flow)
99
+
100
+ ```
101
+ 1. analyzer → Analisa impacto
102
+ 2. ui-ux-reviewer → Pesquisa competidores (se UI)
103
+ 3. documenter → Documenta fluxo
104
+ 4. [IMPLEMENTACAO]
105
+ 5. tester → Cria testes
106
+ 6. security-auditor → Valida seguranca
107
+ 7. quality-checker → typecheck + lint + build
108
+ 8. final-validator → Checklist final
109
+ 9. [COMMIT]
110
+ ```
111
+
112
+ ### Bug Fix (bug_fix_flow)
113
+
114
+ ```
115
+ 1. analyzer → Identifica causa raiz
116
+ 2. [IMPLEMENTACAO]
117
+ 3. tester → Cria teste de regressao
118
+ 4. security-auditor → Valida que fix nao introduz vuln
119
+ 5. quality-checker → typecheck + lint + build
120
+ 6. final-validator → Checklist adaptado
121
+ 7. [COMMIT]
122
+ ```
123
+
124
+ ### Refactor (refactor_flow)
125
+
126
+ ```
127
+ 1. analyzer → Mapeia dependencias
128
+ 2. [IMPLEMENTACAO]
129
+ 3. tester → Atualiza testes
130
+ 4. security-auditor → Valida seguranca
131
+ 5. quality-checker → typecheck + lint + build
132
+ 6. final-validator → Checklist adaptado
133
+ 7. [COMMIT]
134
+ ```
135
+
136
+ ### Config/Chore (config_flow)
137
+
138
+ ```
139
+ 1. quality-checker → Apenas qualidade
140
+ 2. [COMMIT]
141
+ ```
142
+
143
+ ---
144
+
145
+ ## Comandos
146
+
147
+ | Comando | Descricao | Uso |
148
+ | ----------- | ------------------- | ------------------------- |
149
+ | `/feature` | Inicia nova feature | `/feature [descricao]` |
150
+ | `/fix` | Corrige bug | `/fix [descricao do bug]` |
151
+ | `/validate` | Validacao completa | `/validate` |
152
+ | `/research` | Pesquisa topico | `/research [topico]` |
153
+
154
+ ---
155
+
156
+ ## Hooks
157
+
158
+ ### security-check.js
159
+
160
+ Executado ANTES de cada tool call. Bloqueia:
161
+
162
+ - Comandos destrutivos (`rm -rf`, `sudo`, etc)
163
+ - Codigo inseguro (`eval()`, `innerHTML`, etc)
164
+ - Dados sensiveis hardcoded (passwords, API keys)
165
+ - Violacoes de regras do projeto (userId do request, z.any(), etc)
166
+
167
+ ---
168
+
169
+ ## Como Usar
170
+
171
+ ### No Claude Code (Interativo)
172
+
173
+ Os agentes sao carregados automaticamente quando voce abre o projeto.
174
+ O orchestrator coordena o fluxo baseado no tipo de tarefa.
175
+
176
+ ### Com Claude Agent SDK (Programatico)
177
+
178
+ ```python
179
+ from claude_agent_sdk import ClaudeSDKClient
180
+
181
+ client = ClaudeSDKClient(
182
+ setting_sources=["project"] # Carrega .claude/settings.json
183
+ )
184
+
185
+ result = await client.query("Implementar feature X")
186
+ ```
187
+
188
+ ---
189
+
190
+ ## Regras Garantidas pelos Agentes
191
+
192
+ 1. **Seguranca** - security-auditor valida TODAS as regras
193
+ 2. **Testes** - tester garante cobertura adequada
194
+ 3. **Documentacao** - documenter cria docs obrigatorias
195
+ 4. **UI/UX** - ui-ux-reviewer pesquisa competidores
196
+ 5. **Qualidade** - quality-checker roda verificacoes
197
+ 6. **Conformidade** - final-validator valida TUDO antes do commit
198
+
199
+ ---
200
+
201
+ ## Atualizacao
202
+
203
+ Ao adicionar novas regras ao projeto:
204
+
205
+ 1. Atualizar `/claude.md` com a regra
206
+ 2. Atualizar agente relevante em `.claude/agents/`
207
+ 3. Atualizar `final-validator.md` com novo item no checklist
208
+ 4. Se necessario, adicionar pattern em `hooks/security-check.js`
@@ -14,7 +14,7 @@ You manage commits, merges, and are the FINAL agent in the workflow.
14
14
  ## Workflow Order
15
15
 
16
16
  ```
17
- final-validator -> domain-updater -> commit-manager
17
+ quality gates -> commit-manager
18
18
  ```
19
19
 
20
20
  ## Pre-Commit Checklist
@@ -1,32 +1,27 @@
1
- # Project Configuration
2
-
3
- This directory contains project-specific configuration files that agents and skills read to adapt their behavior to your project.
4
-
5
- ## Files
6
-
7
- | File | Purpose |
8
- | --------------------- | ------------------------------------------------------- |
9
- | `project-config.json` | Main project configuration (stack, structure, commands) |
10
- | `domain-mapping.json` | Maps file patterns to knowledge domains |
11
- | `quality-gates.json` | Quality check commands |
12
- | `testing-config.json` | Testing framework and conventions |
13
- | `security-rules.json` | Security audit rules |
14
-
15
- ## How Agents and Skills Use These
16
-
17
- 1. **domain-updater** agent - Reads `domain-mapping.json` for file-to-domain mapping
18
- 2. **documenter** agent - Reads `domain-mapping.json` and `project-config.json`
19
- 3. **tester-unit** agent - Reads `testing-config.json` for frameworks and conventions
20
- 4. **commit-manager** agent - Reads `project-config.json` for commit conventions
21
- 5. **quality-gate** skill - Reads `quality-gates.json` for commands
22
- 6. **security-scan** skill - Reads `security-rules.json` for security patterns
23
- 7. **codebase-knowledge** skill - Reads `domain-mapping.json` for domain structure
24
-
25
- ## Customizing for Your Project
26
-
27
- 1. Update `project-config.json` with your stack and commands
28
- 2. Update `domain-mapping.json` with your project structure
29
- 3. Update `testing-config.json` with your test conventions
30
- 4. Update `security-rules.json` with your auth framework
31
-
32
- The agents and skills will automatically adapt to your configuration.
1
+ # Project Configuration
2
+
3
+ This directory contains project-specific configuration files that agents and skills read to adapt their behavior to your project.
4
+
5
+ ## Files
6
+
7
+ | File | Purpose |
8
+ | --------------------- | ------------------------------------------------------- |
9
+ | `project-config.json` | Main project configuration (stack, structure, commands) |
10
+ | `quality-gates.json` | Quality check commands |
11
+ | `testing-config.json` | Testing framework and conventions |
12
+ | `security-rules.json` | Security audit rules |
13
+
14
+ ## How Agents and Skills Use These
15
+
16
+ 1. **tester-unit** agent - Reads `testing-config.json` for frameworks and conventions
17
+ 2. **commit-manager** agent - Reads `project-config.json` for commit conventions
18
+ 3. **quality-gate** skill - Reads `quality-gates.json` for commands
19
+ 4. **security-scan** skill - Reads `security-rules.json` for security patterns
20
+
21
+ ## Customizing for Your Project
22
+
23
+ 1. Update `project-config.json` with your stack and commands
24
+ 2. Update `testing-config.json` with your test conventions
25
+ 3. Update `security-rules.json` with your auth framework
26
+
27
+ The agents and skills will automatically adapt to your configuration.
@@ -1,70 +1,74 @@
1
- {
2
- "model": "opus",
3
- "max_tokens": 8192,
4
- "max_turns": 100,
5
-
6
- "attribution": {
7
- "commit": "",
8
- "pr": ""
9
- },
10
-
11
- "enableAllProjectMcpServers": true,
12
-
13
- "extraKnownMarketplaces": {
14
- "claude-plugins-official": {
15
- "source": {
16
- "source": "github",
17
- "repo": "anthropics/claude-plugins-official"
18
- }
19
- }
20
- },
21
-
22
- "enabledPlugins": {
23
- "typescript-lsp@claude-plugins-official": true,
24
- "code-review@claude-plugins-official": true,
25
- "security-guidance@claude-plugins-official": true,
26
- "commit-commands@claude-plugins-official": true,
27
- "frontend-design@claude-plugins-official": true
28
- },
29
-
30
- "permissions": {
31
- "mode": "strict",
32
- "allowed_tools": [
33
- "file_read",
34
- "file_write",
35
- "file_edit",
36
- "bash_grep",
37
- "bash_find",
38
- "bash_ls",
39
- "bash_bun",
40
- "bash_git",
41
- "web_search"
42
- ],
43
- "disallowed_tools": [
44
- "bash_rm_rf",
45
- "bash_sudo",
46
- "bash_chmod",
47
- "bash_chown",
48
- "file_delete_recursive"
49
- ],
50
- "allowed_directories": [
51
- ".",
52
- "app",
53
- "components",
54
- "lib",
55
- "server",
56
- "tests",
57
- "docs",
58
- "types",
59
- ".claude"
60
- ],
61
- "protected_files": [".env", ".env.local", ".env.production", "bun.lockb"],
62
- "allow": [
63
- "Bash(bun run typecheck:*)",
64
- "Bash(git add:*)",
65
- "Bash(git commit:*)",
66
- "Bash(bun .claude/scripts/*)",
67
- "Bash(claude mcp:*)"
68
- ]
69
- }
70
- }
1
+ {
2
+ "model": "opus",
3
+ "max_tokens": 8192,
4
+ "max_turns": 100,
5
+
6
+ "attribution": {
7
+ "commit": "",
8
+ "pr": ""
9
+ },
10
+
11
+ "enableAllProjectMcpServers": true,
12
+
13
+ "extraKnownMarketplaces": {
14
+ "claude-plugins-official": {
15
+ "source": {
16
+ "source": "github",
17
+ "repo": "anthropics/claude-plugins-official"
18
+ }
19
+ }
20
+ },
21
+
22
+ "enabledPlugins": {
23
+ "typescript-lsp@claude-plugins-official": true,
24
+ "security-guidance@claude-plugins-official": true,
25
+ "code-review@claude-plugins-official": true,
26
+ "commit-commands@claude-plugins-official": true,
27
+ "frontend-design@claude-plugins-official": true,
28
+ "superpowers@claude-plugins-official": true,
29
+ "ralph-loop@claude-plugins-official": true,
30
+ "context7@claude-plugins-official": true,
31
+ "code-simplifier@claude-plugins-official": true
32
+ },
33
+
34
+ "permissions": {
35
+ "mode": "strict",
36
+ "allowed_tools": [
37
+ "file_read",
38
+ "file_write",
39
+ "file_edit",
40
+ "bash_grep",
41
+ "bash_find",
42
+ "bash_ls",
43
+ "bash_bun",
44
+ "bash_git",
45
+ "web_search"
46
+ ],
47
+ "disallowed_tools": [
48
+ "bash_rm_rf",
49
+ "bash_sudo",
50
+ "bash_chmod",
51
+ "bash_chown",
52
+ "file_delete_recursive"
53
+ ],
54
+ "allowed_directories": [
55
+ ".",
56
+ "app",
57
+ "components",
58
+ "lib",
59
+ "server",
60
+ "tests",
61
+ "docs",
62
+ "types",
63
+ ".claude"
64
+ ],
65
+ "protected_files": [".env", ".env.local", ".env.production", "bun.lockb"],
66
+ "allow": [
67
+ "Bash(bun run typecheck:*)",
68
+ "Bash(git add:*)",
69
+ "Bash(git commit:*)",
70
+ "Bash(bun .claude/scripts/*)",
71
+ "Bash(claude mcp:*)"
72
+ ]
73
+ }
74
+ }