@ofelixdev/cc-kit 1.0.0 → 2.0.0

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.
Files changed (3) hide show
  1. package/README.md +103 -36
  2. package/bin/index.js +31 -0
  3. package/package.json +1 -1
package/README.md CHANGED
@@ -1,8 +1,8 @@
1
1
  # cc-kit
2
2
 
3
- CLI que instala uma knowledge base completa no Claude Code — agents, skills, workflows, scripts de validacao.
3
+ CLI que instala uma knowledge base completa no Claude Code — agents, skills, workflows, hooks, rules e scripts de validacao.
4
4
 
5
- Um comando e seu Claude Code ganha **20 agents especialistas**, **37+ skills**, **11 workflows** e **scripts Python de validacao**. Tudo vai pra `.claude/` do seu projeto.
5
+ Um comando e seu Claude Code ganha **20 agents especialistas**, **108 skills**, **11 workflows**, **4 hooks de automacao**, **6 rules contextuais** e **scripts Python de validacao**. Tudo vai pra `.claude/` do seu projeto.
6
6
 
7
7
  ---
8
8
 
@@ -10,9 +10,9 @@ Um comando e seu Claude Code ganha **20 agents especialistas**, **37+ skills**,
10
10
 
11
11
  O Claude Code por padrao nao sabe nada sobre seu projeto. Ele e poderoso, mas generico.
12
12
 
13
- O **cc-kit** resolve isso instalando uma base de conhecimento curada dentro do `.claude/` — agents que funcionam como especialistas (backend, frontend, seguranca, mobile, etc), skills com patterns e boas praticas, workflows pra tarefas estruturadas, e scripts Python pra validacao automatica.
13
+ O **cc-kit** resolve isso instalando uma base de conhecimento curada dentro do `.claude/` — agents que funcionam como especialistas (backend, frontend, seguranca, mobile, etc), skills com patterns e boas praticas, workflows pra tarefas estruturadas, hooks que automatizam formatacao e protecao, rules que ativam automaticamente por contexto, e scripts Python pra validacao.
14
14
 
15
- Basicamente: **seu Claude Code vira um time de especialistas**.
15
+ Basicamente: **seu Claude Code vira um time de especialistas que age automaticamente**.
16
16
 
17
17
  ---
18
18
 
@@ -35,16 +35,17 @@ cc-kit init
35
35
 
36
36
  ```
37
37
  .claude/
38
- ├── agents/ # 20 agents especialistas (.md)
39
- ├── skills/ # 37+ skills com patterns e boas praticas
40
- ├── workflows/ # 11 workflows estruturados
41
- ├── scripts/ # Scripts Python de validacao (checklist, verify_all, etc)
42
- ├── rules/CLAUDE.md # Regras estendidas de coding standards
43
- ├── .shared/ # Dados compartilhados (UI/UX datasets, etc)
44
- ├── ARCHITECTURE.md # Indice completo da knowledge base
45
- └── mcp_config.json # Config MCP servers (com placeholder API key)
46
-
47
- CLAUDE.md # Regras do projeto na raiz (criado se nao existir)
38
+ ├── agents/ # 20 agents especialistas
39
+ ├── skills/ # 108 skills (cc-kit + trailofbits + vercel + community)
40
+ ├── workflows/ # 11 workflows estruturados
41
+ ├── hooks/ # 4 hooks de automacao (format, protect, notify, compact)
42
+ ├── rules/ # 6 rules (1 global + 5 path-conditional)
43
+ ├── scripts/ # Scripts Python de validacao
44
+ ├── settings.template.json → settings.json (criado na primeira instalacao)
45
+ ├── ARCHITECTURE.md # Indice completo da knowledge base
46
+ └── mcp_config.json # Config MCP servers
47
+
48
+ CLAUDE.md # Regras do projeto na raiz (criado se nao existir)
48
49
  ```
49
50
 
50
51
  ---
@@ -56,7 +57,7 @@ CLAUDE.md # Regras do projeto na raiz (criado se nao existir)
56
57
  Baixa e instala a knowledge base no `.claude/`.
57
58
 
58
59
  - **Merge inteligente**: se `.claude/` ja existe, faz merge — seu conteudo existente e preservado
59
- - **Paths protegidos**: `settings.json`, `settings.local.json`, `plans/`, `memory/`, `projects/` **nunca** sao tocados
60
+ - **Paths protegidos**: `settings.json`, `settings.local.json`, `plans/`, `memory/`, `projects/` **nunca** sao sobrescritos
60
61
  - `CLAUDE.md` da raiz so e criado se nao existir (use `--force` pra sobrescrever)
61
62
 
62
63
  ```bash
@@ -86,33 +87,99 @@ cc-kit status
86
87
 
87
88
  ---
88
89
 
89
- ## O que vem dentro
90
+ ## Principais Features
90
91
 
91
- ### Agents (20)
92
+ ### Sistema Autonomo de Agents
92
93
 
93
- Arquivos de conhecimento especialista que o Claude le on-demand:
94
+ O `CLAUDE.md` instalado na raiz funciona como diretiva — o Claude Code automaticamente detecta o dominio da tarefa e carrega o agent + skills correspondentes. Sem precisar pedir nada.
94
95
 
95
- `orchestrator` · `project-planner` · `security-auditor` · `backend-specialist` · `frontend-specialist` · `mobile-developer` · `debugger` · `game-developer` · `test-engineer` · `database-architect` · `devops-engineer` · `performance-optimizer` · `seo-specialist` · `penetration-tester` · `documentation-writer` · `qa-automation-engineer` · `code-archaeologist` · `explorer-agent` · `product-manager` · `product-owner`
96
+ | Dominio | Agent | Skills |
97
+ |---|---|---|
98
+ | Frontend, React, UI | `frontend-specialist` | frontend-design, react-best-practices, tailwind-patterns |
99
+ | Backend, API, Node | `backend-specialist` | api-patterns, nodejs-best-practices |
100
+ | Database, Schema | `database-architect` | database-design |
101
+ | Seguranca, Auth | `security-auditor` | vulnerability-scanner, codeql, semgrep |
102
+ | Testes, Coverage | `test-engineer` | testing-patterns, webapp-testing, coverage-analysis |
103
+ | Debug, Bugs | `debugger` | systematic-debugging |
104
+ | DevOps, Deploy | `devops-engineer` | deployment-procedures, devops, coolify |
105
+ | Mobile | `mobile-developer` | mobile-design |
106
+ | Games | `game-developer` | game-development |
107
+ | Performance | `performance-optimizer` | performance-profiling |
108
+ | SEO | `seo-specialist` | seo-fundamentals |
109
+ | Novo projeto | `orchestrator` | app-builder, architecture |
96
110
 
97
- ### Skills (37+)
111
+ ### Hooks de Automacao (4)
98
112
 
99
- Patterns, boas praticas e conhecimento por dominio:
113
+ Hooks pre-configurados em `settings.json`:
100
114
 
101
- `clean-code` · `api-patterns` · `database-design` · `frontend-design` · `mobile-design` · `testing-patterns` · `vulnerability-scanner` · `brainstorming` · `plan-writing` · `architecture` · `tailwind-patterns` · `nextjs-react-expert` · `nodejs-best-practices` · `python-patterns` · `rust-pro` · `bash-linux` · `game-development` · `seo-fundamentals` · `performance-profiling` · `webapp-testing` · `tdd-workflow` · `code-review-checklist` · `deployment-procedures` · `mcp-builder` · e mais...
115
+ | Hook | Evento | O que faz |
116
+ |---|---|---|
117
+ | `auto-format.sh` | Apos editar arquivos | Formata com prettier/eslint automaticamente |
118
+ | `protect-files.sh` | Antes de editar | Bloqueia escrita em `.env`, secrets, credentials |
119
+ | `notify-done.sh` | Ao terminar | Notificacao desktop (macOS/Linux) |
120
+ | `post-compact.sh` | Apos compactacao | Re-injeta contexto do projeto |
102
121
 
103
- ### Workflows (11)
122
+ ### Rules Path-Conditional (5)
104
123
 
105
- Fluxos estruturados de execucao de tarefas:
124
+ Rules em `.claude/rules/` que ativam automaticamente baseado no arquivo sendo editado:
106
125
 
107
- `brainstorm` · `create` · `debug` · `deploy` · `enhance` · `orchestrate` · `plan` · `preview` · `status` · `test` · `ui-ux-pro-max`
126
+ | Rule | Ativa para | O que aplica |
127
+ |---|---|---|
128
+ | `frontend.md` | `*.tsx`, `*.jsx`, `components/`, `pages/` | Acessibilidade, performance, component patterns |
129
+ | `backend.md` | `api/`, `services/`, `controllers/` | Input validation, error handling, seguranca |
130
+ | `testing.md` | `*.test.*`, `__tests__/` | Padrao AAA, isolamento, testes por comportamento |
131
+ | `security.md` | `auth/`, `middleware/`, `session*` | Sem hardcoded secrets, hashing, CSRF, rate limiting |
132
+ | `database.md` | `prisma/`, `migrations/`, `models/` | Migrations, indexes, transactions, soft deletes |
108
133
 
109
- ### Scripts de Validacao
134
+ ### Settings Template
110
135
 
111
- Scripts Python pra rodar checks no seu projeto:
136
+ Na primeira instalacao, cria um `settings.json` com:
137
+ - Permissoes deny para `.env` e secrets
138
+ - Hooks pre-configurados (format, protect, notify, compact)
139
+ - Nunca sobrescreve se ja existir
140
+
141
+ ---
142
+
143
+ ## Skills (108)
144
+
145
+ ### Por Fonte
146
+
147
+ | Fonte | Qtd | Foco |
148
+ |---|---|---|
149
+ | **cc-kit** | 36 | Full-stack development |
150
+ | **trailofbits/skills** | 60 | Seguranca, fuzzing, auditoria |
151
+ | **vercel-labs** | 2 | Web design, skill discovery |
152
+ | **obra/superpowers** | 1 | Brainstorming colaborativo |
153
+ | **evolv3-ai/vibe-skills** | 3 | Coolify, DevOps, cloud infra |
154
+ | **addyosmani/web-quality** | 6 | Acessibilidade, performance, SEO |
155
+
156
+ ### Por Categoria
157
+
158
+ **Frontend & UI**: react-best-practices, web-design-guidelines, tailwind-patterns, frontend-design, ui-ux-pro-max, nextjs-react-expert
159
+
160
+ **Backend & API**: api-patterns, nodejs-best-practices, python-patterns, modern-python
161
+
162
+ **Database**: database-design
163
+
164
+ **Cloud & Infra**: deployment-procedures, server-management, devcontainer-setup, coolify, coolify-cli, devops
165
+
166
+ **Testing & Quality**: testing-patterns, webapp-testing, tdd-workflow, code-review-checklist, lint-and-validate, coverage-analysis, property-based-testing, debug-buttercup, differential-review
167
+
168
+ **Seguranca**: vulnerability-scanner, red-team-tactics, codeql, semgrep, entry-point-analyzer, supply-chain-risk-auditor, e 30+ mais
169
+
170
+ **Fuzzing**: aflpp, libfuzzer, cargo-fuzz, atheris, harness-writing, libafl, ossfuzz, e mais
171
+
172
+ **Blockchain**: algorand, cairo, cosmos, solana, substrate, ton vulnerability scanners
173
+
174
+ **Arquitetura**: app-builder, architecture, plan-writing, brainstorming
175
+
176
+ ---
177
+
178
+ ## Scripts de Validacao
112
179
 
113
180
  ```bash
114
- python .claude/scripts/checklist.py . # Validacao por prioridade
115
- python .claude/scripts/verify_all.py . --url <URL> # Suite completa pre-deploy
181
+ python .claude/scripts/checklist.py . # Quick: Security → Lint → Schema → Tests → UX → SEO
182
+ python .claude/scripts/verify_all.py . --url <URL> # Full: + Lighthouse, E2E, Mobile, i18n
116
183
  ```
117
184
 
118
185
  ---
@@ -120,9 +187,11 @@ python .claude/scripts/verify_all.py . --url <URL> # Suite completa pre-deploy
120
187
  ## Como funciona
121
188
 
122
189
  1. Baixa o template do GitHub via [giget](https://github.com/unjs/giget)
123
- 2. Faz merge dos agents, skills, workflows e scripts no `.claude/`
124
- 3. Cria um `CLAUDE.md` generico na raiz do projeto (se nao existir)
125
- 4. Nunca toca em configs do usuario (settings, memory, plans)
190
+ 2. Faz merge dos agents, skills, workflows, scripts, hooks e rules no `.claude/`
191
+ 3. Cria `settings.json` com hooks pre-configurados (se nao existir)
192
+ 4. Cria `CLAUDE.md` na raiz do projeto (se nao existir)
193
+ 5. Torna hooks executaveis (`chmod +x`)
194
+ 6. Nunca toca em configs do usuario (settings existente, memory, plans)
126
195
 
127
196
  ---
128
197
 
@@ -136,11 +205,9 @@ python .claude/scripts/verify_all.py . --url <URL> # Suite completa pre-deploy
136
205
 
137
206
  ## Creditos
138
207
 
139
- Este projeto foi inspirado no [`@vudovn/ag-kit`](https://www.npmjs.com/package/@vudovn/ag-kit) (Antigravity Kit), que instala agents/skills/workflows pro Gemini CLI no diretorio `.agent/`.
140
-
141
- O **cc-kit** adapta esse mesmo conceito pro ecossistema **Claude Code** — todo o conteudo foi reescrito e adaptado pra funcionar com a estrutura `.claude/`, sem nenhuma referencia ao Gemini ou Antigravity.
208
+ Inspirado no [`@vudovn/ag-kit`](https://www.npmjs.com/package/@vudovn/ag-kit) (Antigravity Kit). Obrigado ao [@vudovn](https://github.com/vudovn) pela ideia original.
142
209
 
143
- Obrigado ao [@vudovn](https://github.com/vudovn) pela ideia original e pelo trabalho no ag-kit.
210
+ Skills de seguranca por [Trail of Bits](https://github.com/trailofbits/skills). Skills de web quality por [Addy Osmani](https://github.com/nicepkg/claude-code-awesome-skills). Brainstorming por [obra/superpowers](https://github.com/obra/superpowers). Coolify/DevOps por [evolv3-ai](https://github.com/evolv3-ai/vibe-skills).
144
211
 
145
212
  ---
146
213
 
package/bin/index.js CHANGED
@@ -30,6 +30,7 @@ const MERGE_DIRS = [
30
30
  'workflows',
31
31
  'scripts',
32
32
  'rules',
33
+ 'hooks',
33
34
  '.shared',
34
35
  ];
35
36
 
@@ -140,6 +141,33 @@ async function installTemplate({ force, targetPath, branch, quiet, dryRun }) {
140
141
  console.log(chalk.dim(` ${ROOT_CLAUDE_MD} already exists — skipped (use --force to overwrite)`));
141
142
  }
142
143
 
144
+ // Install settings.json from template if not exists (never overwrite)
145
+ const settingsPath = path.join(claude, 'settings.json');
146
+ const settingsExists = await fs.pathExists(settingsPath);
147
+ if (!settingsExists) {
148
+ const settingsTemplate = path.join(tempDir, 'settings.template.json');
149
+ if (await fs.pathExists(settingsTemplate)) {
150
+ await fs.copy(settingsTemplate, settingsPath);
151
+ totalCopied++;
152
+ if (!quiet) {
153
+ console.log(chalk.dim(' Created settings.json with hooks & permissions'));
154
+ }
155
+ }
156
+ } else if (!quiet) {
157
+ console.log(chalk.dim(' settings.json already exists — skipped'));
158
+ }
159
+
160
+ // Make hook scripts executable
161
+ const hooksDir = path.join(claude, 'hooks');
162
+ if (await fs.pathExists(hooksDir)) {
163
+ const hookFiles = await fs.readdir(hooksDir);
164
+ for (const file of hookFiles) {
165
+ if (file.endsWith('.sh')) {
166
+ await fs.chmod(path.join(hooksDir, file), 0o755);
167
+ }
168
+ }
169
+ }
170
+
143
171
  if (spinner) spinner.succeed(chalk.green(`Installed ${totalCopied} files into ${TARGET_DIR}/`));
144
172
 
145
173
  if (!quiet) {
@@ -148,6 +176,8 @@ async function installTemplate({ force, targetPath, branch, quiet, dryRun }) {
148
176
  console.log(chalk.dim(' Skills: ') + await countItems(path.join(claude, 'skills')));
149
177
  console.log(chalk.dim(' Workflows: ') + await countItems(path.join(claude, 'workflows')));
150
178
  console.log(chalk.dim(' Scripts: ') + await countItems(path.join(claude, 'scripts')));
179
+ console.log(chalk.dim(' Hooks: ') + await countItems(path.join(claude, 'hooks')));
180
+ console.log(chalk.dim(' Rules: ') + await countItems(path.join(claude, 'rules')));
151
181
  console.log('');
152
182
  console.log(chalk.green(' Done!') + chalk.dim(' Run ') + chalk.cyan('cc-kit status') + chalk.dim(' to verify.'));
153
183
  console.log('');
@@ -197,6 +227,7 @@ async function showStatus(targetPath) {
197
227
  console.log(chalk.dim(' Workflows: ') + await countItems(path.join(claude, 'workflows')));
198
228
  console.log(chalk.dim(' Scripts: ') + await countItems(path.join(claude, 'scripts')));
199
229
  console.log(chalk.dim(' Rules: ') + await countItems(path.join(claude, 'rules')));
230
+ console.log(chalk.dim(' Hooks: ') + await countItems(path.join(claude, 'hooks')));
200
231
  console.log('');
201
232
  }
202
233
 
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@ofelixdev/cc-kit",
3
- "version": "1.0.0",
3
+ "version": "2.0.0",
4
4
  "type": "module",
5
5
  "description": "Claude Code knowledge base installer - agents, skills, workflows",
6
6
  "main": "bin/index.js",