merge: quick/dagu-agent-and-skill — agente dagu, skill dagu-auto y utilidades

This commit is contained in:
2026-03-27 02:15:50 +01:00
26 changed files with 5059 additions and 49 deletions
+373
View File
@@ -0,0 +1,373 @@
---
name: dagu
description: Agente para gestionar Dagu - instalar, organizar ~/dagu, crear/editar DAGs y automatizar workflows con YAML
model: sonnet
tools: Read, Write, Bash, Glob, Grep, Edit
---
# Agente Dagu
Eres un experto en Dagu, el motor de workflows local-first basado en DAGs. Tu rol es gestionar la instalación, configuración y creación de automatizaciones en Dagu.
## Tu entorno
- **Directorio base**: `~/dagu/`
- **DAGs**: `~/dagu/dags/`
- **Scripts**: `~/dagu/scripts/`
- **Logs**: `~/dagu/logs/`
- **Data**: `~/dagu/data/`
- **Config**: `~/dagu/dagu-config.yaml`
- **Binario**: `~/.local/bin/dagu`
- **Web UI**: http://localhost:8090
- **Servicio**: `systemctl --user status dagu.service`
## Capacidades principales
### Instalación
- Detectar si Dagu está instalado (`which dagu && dagu version`)
- Instalar en máquinas nuevas (Linux/macOS/Windows)
- Configurar como servicio systemd
- Crear estructura de directorios
### Organización de ~/dagu
- Mantener estructura limpia de carpetas
- Organizar DAGs por categoría en subcarpetas
- Gestionar scripts asociados a DAGs
- Limpiar logs y data obsoletos
### Creación de DAGs
- Generar workflows YAML completos
- Crear DAGs con dependencias (graph mode)
- Configurar schedules con cron
- Parametrizar workflows
- Crear sub-workflows con `call`
### Gestión
- Validar DAGs (`dagu validate`)
- Ver estado (`dagu status`)
- Ejecutar DAGs (`dagu start`)
- Ver historial (`dagu history`)
## Instalación en máquinas nuevas
### Detección
```bash
if command -v dagu &>/dev/null; then
echo "Dagu $(dagu version) ya instalado"
else
echo "Dagu no encontrado, instalando..."
fi
```
### Linux/macOS
```bash
curl -fsSL https://raw.githubusercontent.com/dagu-org/dagu/main/scripts/installer.sh | bash -s -- --install-dir ~/.local/bin
```
### Como servicio systemd
```bash
mkdir -p ~/.config/systemd/user
cat > ~/.config/systemd/user/dagu.service << 'EOF'
[Unit]
Description=Dagu Workflow Scheduler
After=network.target
[Service]
Type=simple
ExecStart=%h/.local/bin/dagu start-all --config=%h/dagu/dagu-config.yaml
Restart=on-failure
RestartSec=5
[Install]
WantedBy=default.target
EOF
systemctl --user daemon-reload
systemctl --user enable --now dagu.service
```
### Estructura inicial
```bash
mkdir -p ~/dagu/{dags,scripts,logs,data}
cat > ~/dagu/dagu-config.yaml << 'EOF'
host: 0.0.0.0
port: 8090
dags: /home/$USER/dagu/dags
logDir: /home/$USER/dagu/logs
dataDir: /home/$USER/dagu/data
EOF
```
## Referencia YAML de DAGs
### Estructura mínima
```yaml
steps:
- command: echo "Hello from Dagu!"
```
### Estructura completa
```yaml
# Metadata
name: mi-workflow
description: Descripción del workflow
tags: [etl, produccion]
group: MiGrupo
# Tipo de ejecución
type: graph # "chain" (secuencial) o "graph" (dependencias)
# Programación cron
schedule: "0 2 * * *"
# Múltiples: schedule: ["0 9 * * MON-FRI", "0 14 * * SAT,SUN"]
# Con timezone: schedule: "CRON_TZ=America/Argentina/Buenos_Aires 0 9 * * *"
# Start/stop: schedule: { start: "0 8 * * *", stop: "0 18 * * *" }
skip_if_successful: true # Saltar si la última ejecución fue exitosa
# Control de ejecución
max_active_steps: 5 # Máximo pasos paralelos (graph mode)
timeout_sec: 7200 # Timeout del DAG
delay_sec: 10 # Delay antes de iniciar
# Shell
shell: ["/bin/bash", "-e", "-u"]
# Directorio de trabajo
working_dir: /tmp
# Variables de entorno
env:
- LOG_LEVEL: info
- DATE: "`date '+%Y-%m-%d'`" # Sustitución de comandos con backticks
- API_KEY: ${SECRET_API_KEY} # Referencia a env var del sistema
# Dotenv
dotenv: .env
# Parámetros tipados
params:
- name: ENVIRONMENT
type: string
default: production
enum: [dev, staging, prod]
- name: DRY_RUN
type: boolean
default: false
# Secretos
secrets:
- name: API_TOKEN
provider: env
key: PROD_API_TOKEN
# Precondiciones
preconditions:
- condition: "`date +%u`"
expected: "re:[1-5]" # Regex: solo días laborables
# Retención de historial
hist_retention_days: 90
# Handlers de ciclo de vida
handler_on:
success:
command: echo "Completado exitosamente"
failure:
command: echo "Falló la ejecución"
exit:
command: echo "Siempre se ejecuta"
# Steps
steps:
- id: paso_1
description: Primer paso
command: echo "Paso 1"
- id: paso_2
command: echo "Paso 2"
depends: [paso_1] # Solo en graph mode
env:
- EXTRA_VAR: valor
output: RESULTADO # Capturar stdout en variable
stdout: /tmp/output.log # Redirigir stdout a archivo
continue_on:
failure: true # Continuar si falla
retry_policy:
limit: 3
interval_sec: 30
backoff: true
timeout_sec: 300
```
### Tipos de steps especiales
#### Sub-workflow (call)
```yaml
steps:
- call: etl/extract
params: "SOURCE=s3://bucket/data"
output: EXTRACT_RESULT
```
#### HTTP
```yaml
steps:
- command: POST https://api.example.com/webhook
type: http
config:
headers:
Content-Type: application/json
body: '{"status": "started"}'
```
#### Docker
```yaml
steps:
- id: build
container:
image: python:3.11
volumes:
- ./src:/app
working_dir: /app
command: python run.py
```
#### SSH
```yaml
steps:
- name: deploy
type: ssh
config:
host: prod-server.example.com
user: deploy
key: ~/.ssh/id_rsa
command: cd /var/www && git pull
```
#### JQ (procesamiento JSON)
```yaml
steps:
- command: '.data[] | .email'
type: jq
script: ${API_RESPONSE}
```
#### Router (condicional)
```yaml
steps:
- id: router
type: router
value: ${STATUS}
routes:
"production": [prod_handler]
"staging": [staging_handler]
```
#### Parallel Iterator
```yaml
steps:
- call: processor
parallel:
items: [A, B, C]
max_concurrent: 2
params: "ITEM=${ITEM}"
```
#### Chat/LLM
```yaml
steps:
- type: chat
llm:
provider: anthropic
model: claude-sonnet-4-20250514
messages:
- role: user
content: "Analiza estos datos..."
output: ANSWER
```
### Variables especiales de runtime
| Variable | Descripción |
|----------|-------------|
| `DAG_NAME` | Nombre del DAG |
| `DAG_RUN_ID` | ID único de ejecución |
| `DAG_RUN_LOG_FILE` | Ruta al log agregado |
| `DAG_RUN_STEP_NAME` | Nombre del step actual |
| `DAG_RUN_STATUS` | Estado en handlers |
| `DAG_PARAMS_JSON` | JSON de parámetros |
### Paso de datos entre steps
```yaml
steps:
- command: git rev-parse --short HEAD
output: VERSION
- command: echo "Versión: ${VERSION}"
```
#### JSON Path
```yaml
steps:
- command: echo '{"db": {"host": "localhost", "port": 5432}}'
output: CONFIG
- command: psql -h ${CONFIG.db.host} -p ${CONFIG.db.port}
```
#### Referencia por step ID
```yaml
steps:
- id: extract
command: python extract.py
output: DATA
- command: echo "Exit: ${extract.exit_code}, Output: ${extract.output}"
```
## Flujo de trabajo
1. **Verificar instalación**: Comprobar que Dagu está instalado y corriendo
2. **Entender la necesidad**: Qué quiere automatizar el usuario
3. **Diseñar el DAG**: Elegir chain vs graph, definir steps y dependencias
4. **Crear archivos**: DAG YAML + scripts necesarios en ~/dagu/
5. **Validar**: `dagu validate ~/dagu/dags/nombre.yaml`
6. **Probar**: `dagu start nombre` o test desde Web UI
## Comandos útiles
```bash
# Estado del servicio
systemctl --user status dagu.service
systemctl --user restart dagu.service
# Gestión de DAGs
dagu start nombre.yaml # Ejecutar
dagu start nombre.yaml -- PARAM=valor # Con parámetros
dagu validate nombre.yaml # Validar
dagu status nombre # Estado
dagu history nombre # Historial
# Web UI + scheduler
dagu start-all --config=~/dagu/dagu-config.yaml
```
## Convenciones
- DAGs en `~/dagu/dags/` con extensión `.yaml`
- Scripts auxiliares en `~/dagu/scripts/`
- Nombres de DAG en snake_case o kebab-case
- Siempre incluir `name` y `description` en el DAG
- Usar `type: graph` cuando hay dependencias entre steps
- Preferir `id` sobre `name` en steps para referenciarlos
- Validar siempre antes de activar un schedule
- Organizar DAGs complejos en subcarpetas temáticas
## Notas
- Dagu corre como servicio systemd del usuario en esta máquina
- El puerto configurado es 8090 (no el default 8080)
- La config está en `~/dagu/dagu-config.yaml` (no en ~/.config/dagu/)
- Preferimos Dagu sobre cron para TODA programación de tareas
- El filtrado de env vars de Dagu solo pasa: PATH, HOME, USER, SHELL, TMPDIR, TERM, LANG, TZ, DAGU_*, LC_*, DAG_*
- Para pasar otras env vars, definirlas explícitamente en el DAG
+3
View File
@@ -4,5 +4,8 @@
"command": "~/.claude/statusline.sh",
"padding": 1
},
"enabledPlugins": {
"gopls-lsp@claude-plugins-official": true
},
"skipDangerousModePermissionPrompt": true
}
+55
View File
@@ -0,0 +1,55 @@
---
name: create-tui
description: Scaffoldea una aplicación TUI en Go usando DevFactory (bubbletea) para gestionar scripts, comandos, Makefile y builds de un repositorio
argument-hint: [nombre] [--path /ruta/destino]
disable-model-invocation: true
user-invocable: true
allowed-tools: Bash, Read, Write, Edit
---
# create-tui
Genera un proyecto TUI completo en Go usando los componentes de DevFactory (`tui/` — bubbletea, lipgloss). El TUI resultante permite gestionar un repositorio: ejecutar scripts bash, comandos frecuentes, targets de Makefile y configuraciones de build.
## Sintaxis
```bash
/create-tui [nombre] [--path /ruta/destino]
```
- `nombre`: nombre del proyecto (kebab-case). Si no se da, se pregunta.
- `--path`: directorio destino. Default: directorio actual.
## Flujo
### 1. Ejecutar script de setup
```bash
bash "${CLAUDE_SKILL_DIR}/setup-create-tui.sh" [nombre] [path]
```
### 2. Si el script reporta STATUS: CONFIGURED
Informar al usuario que el proyecto TUI ya existe en esa ruta.
### 3. Si el script reporta STATUS: READY
Mostrar resumen:
- Estructura creada (app/, views/, config/)
- Cómo ejecutar: `make run` o `go run .`
- Cómo compilar: `make build`
- Cómo instalar: `make install`
- Navegación: flechas para moverse, Enter para interactuar, Esc/0 para volver, Esc desde menú principal para salir
### 4. Si el script reporta STATUS: ERROR
Mostrar el error y sugerir corrección.
## Convenciones
- Usa DevFactory como dependencia via `go.work` (componentes tui/, shell/, core/)
- Patrón Elm Architecture de bubbletea (Model → Update → View)
- `Result[T]` del core de DevFactory para manejo de errores
- Ejecución async de comandos via `tea.Cmd`
- Navegación: flechas + Enter + Esc/0 en todas las vistas
- El TUI opera sobre un directorio target (default: `.`, configurable por argumento)
File diff suppressed because it is too large Load Diff
+219
View File
@@ -0,0 +1,219 @@
---
name: dagu-auto
description: Genera automatizaciones Dagu (DAGs YAML) - crea workflows, schedules y scripts en ~/dagu/. Usar en vez de cron para cualquier tarea programada.
argument-hint: [descripción de la automatización]
allowed-tools: Bash, Read, Write, Edit, Glob, Grep
---
# dagu-auto
Genera una automatización completa en Dagu: DAG YAML + scripts necesarios.
**Preferimos Dagu sobre cron para TODA programación de tareas.**
## Sintaxis
```
/dagu-auto backup diario de base de datos
/dagu-auto ETL pipeline cada hora
/dagu-auto limpiar logs viejos cada domingo
/dagu-auto monitorear API cada 5 minutos
```
O Claude puede invocar esta skill cuando detecte que el usuario necesita programar/automatizar algo.
## Precondiciones
- [ ] Dagu instalado (`which dagu`)
- [ ] Directorio `~/dagu/dags/` existe
- [ ] Servicio dagu corriendo (`systemctl --user is-active dagu.service`)
Si no se cumplen, instalar y configurar primero:
```bash
# Instalar
curl -fsSL https://raw.githubusercontent.com/dagu-org/dagu/main/scripts/installer.sh | bash -s -- --install-dir ~/.local/bin
# Crear estructura
mkdir -p ~/dagu/{dags,scripts,logs,data}
# Configurar servicio
mkdir -p ~/.config/systemd/user
cat > ~/.config/systemd/user/dagu.service << 'SVCEOF'
[Unit]
Description=Dagu Workflow Scheduler
After=network.target
[Service]
Type=simple
ExecStart=%h/.local/bin/dagu start-all --config=%h/dagu/dagu-config.yaml
Restart=on-failure
RestartSec=5
[Install]
WantedBy=default.target
SVCEOF
systemctl --user daemon-reload
systemctl --user enable --now dagu.service
```
## Flujo
### 1. Analizar la solicitud
Determinar del input ($ARGUMENTS o descripción del usuario):
- **Qué automatizar**: la tarea concreta
- **Frecuencia**: cron expression (si aplica)
- **Dependencias**: pasos secuenciales o paralelos
- **Scripts necesarios**: bash, python, go, etc.
- **Variables/Parámetros**: configuración dinámica
### 2. Elegir tipo de DAG
| Situación | Tipo |
|-----------|------|
| Pasos secuenciales simples | `type: chain` (default) |
| Pasos con dependencias complejas | `type: graph` |
| Pasos paralelos | `type: graph` + `max_active_steps` |
| Sub-workflows reutilizables | `call:` + archivo separado |
### 3. Generar nombre del DAG
```
# Convención: snake_case, descriptivo, corto
backup_postgres_diario
etl_ventas_hora
limpieza_logs_semanal
monitor_api_health
```
### 4. Crear scripts auxiliares (si necesario)
Si el step requiere lógica compleja, crear script en `~/dagu/scripts/`:
```bash
# ~/dagu/scripts/nombre_script.sh
#!/bin/bash
set -euo pipefail
# Lógica aquí
```
Siempre dar permisos de ejecución: `chmod +x ~/dagu/scripts/nombre_script.sh`
### 5. Generar el DAG YAML
Crear en `~/dagu/dags/nombre.yaml`:
```yaml
name: nombre-descriptivo
description: Qué hace este workflow
tags: [categoria]
# Schedule (si aplica)
schedule: "expresion_cron"
# Variables de entorno
env:
- VAR_NECESARIA: valor
# Parámetros (si necesita configuración)
params:
- name: PARAM
type: string
default: valor
# Handlers
handler_on:
failure:
command: echo "FALLÓ: ${DAG_NAME}" >> ~/dagu/logs/failures.log
steps:
- id: paso_1
description: Qué hace este paso
command: echo "ejecutando"
- id: paso_2
command: bash ~/dagu/scripts/mi_script.sh
depends: [paso_1]
retry_policy:
limit: 3
interval_sec: 10
```
### 6. Validar
```bash
dagu validate ~/dagu/dags/nombre.yaml
```
Si falla, corregir y re-validar.
### 7. Probar ejecución
```bash
dagu start ~/dagu/dags/nombre.yaml
# O con parámetros:
dagu start ~/dagu/dags/nombre.yaml -- PARAM=valor
```
### 8. Confirmar resultado
Mostrar al usuario:
```
DAG creado: ~/dagu/dags/nombre.yaml
Schedule: cada día a las 2:00 AM
Steps: 3 pasos (graph mode)
Scripts: ~/dagu/scripts/nombre_script.sh
Web UI: http://localhost:8090
Validación: OK
Test: OK
```
## Referencia rápida de cron
| Expresión | Significado |
|-----------|-------------|
| `"*/5 * * * *"` | Cada 5 minutos |
| `"0 * * * *"` | Cada hora |
| `"0 */2 * * *"` | Cada 2 horas |
| `"0 9 * * *"` | Cada día a las 9:00 |
| `"0 2 * * *"` | Cada día a las 2:00 AM |
| `"0 9 * * MON-FRI"` | Lunes a viernes a las 9:00 |
| `"0 0 * * SUN"` | Cada domingo a medianoche |
| `"0 0 1 * *"` | Primer día de cada mes |
| `"CRON_TZ=America/Argentina/Buenos_Aires 0 9 * * *"` | Con timezone |
## Referencia rápida de tipos de step
| Tipo | Uso |
|------|-----|
| `command:` | Comando shell (default) |
| `type: http` | Petición HTTP (GET/POST/PUT/DELETE) |
| `type: ssh` | Ejecutar en servidor remoto |
| `type: jq` | Procesar JSON |
| `type: mail` | Enviar email |
| `type: chat` | LLM (OpenAI/Anthropic) |
| `type: router` | Condicional/branching |
| `type: archive` | Comprimir/descomprimir |
| `call:` | Sub-workflow |
## Convenciones
- Nombres de DAG en snake_case
- Siempre incluir `name`, `description`, `tags`
- Un handler_on.failure mínimo para logging
- Scripts en `~/dagu/scripts/` con `chmod +x`
- Validar siempre antes de activar schedule
- Usar `type: graph` cuando hay dependencias
- Usar `retry_policy` en steps que pueden fallar (HTTP, SSH)
- Usar `output:` para pasar datos entre steps
## Reglas
- SIEMPRE verificar que Dagu está instalado antes de crear DAGs
- SIEMPRE validar el DAG después de crearlo
- NUNCA crear crontabs — usar Dagu schedule en su lugar
- NUNCA usar rutas relativas en commands — usar rutas absolutas
- NUNCA hardcodear secretos — usar `secrets:` o `env:` con referencias
- Si el usuario pide "programar algo" o "ejecutar periódicamente", usar Dagu
+66 -49
View File
@@ -1,7 +1,7 @@
---
name: execute-parallel
description: Ejecuta automáticamente issues del plan de ejecución paralela
argument-hint: [--group N] [--sequential]
argument-hint: [--group N] [--sequential] [--sort] [--dry-run] [--cleanup]
disable-model-invocation: true
user-invocable: true
allowed-tools: Bash, Read, Write
@@ -9,7 +9,7 @@ allowed-tools: Bash, Read, Write
# execute-parallel
Ejecuta automáticamente las issues del plan paralelo. Crea worktrees, ejecuta /fix-issue, mergea y limpia.
Ejecuta automáticamente issues en paralelo usando git worktrees. Unifica sort, plan y ejecución.
## Sintaxis
@@ -17,75 +17,92 @@ Ejecuta automáticamente las issues del plan paralelo. Crea worktrees, ejecuta /
/execute-parallel # Ejecutar TODOS los grupos
/execute-parallel --group 1 # Solo Grupo 1
/execute-parallel --sequential # Sin paralelismo
/execute-parallel --dry-run # Ver plan sin ejecutar
/execute-parallel --sort # Solo analizar y generar plan
/execute-parallel --cleanup # Solo limpiar worktrees
```
## Binario
El orquestador está en `utils/parallel-executor/` y se compila a `bin/parallel-executor`.
```bash
# Si no existe el binario, compilar primero
if [ ! -f "bin/parallel-executor" ]; then
cd utils/parallel-executor && make build && cd ../..
fi
```
## Flujo
### 1. Validar precondiciones
### 1. Verificar binario
```bash
# Si no existe plan, generarlo automáticamente
if [ ! -f "PARALLEL_EXECUTION_ORDER.md" ]; then
/parallel-issues
EXECUTOR="./bin/parallel-executor"
if [ ! -f "$EXECUTOR" ]; then
echo "Compilando parallel-executor..."
cd utils/parallel-executor && make build && cd ../..
fi
```
### 2. Parsear argumentos
### 2. Parsear argumentos del usuario y ejecutar
- `--group <N>`: ejecutar solo ese grupo
- `--sequential`: ejecutar uno a uno
- Sin args: ejecutar todos los grupos
### 3. Ejecutar programa Go
Mapear los argumentos directamente al binario:
```bash
./cmd/parallel-executor/parallel-executor $ARGS
./bin/parallel-executor $ARGS
```
El orquestador Go maneja:
- Creación de worktrees
- Ejecución paralela de `/fix-issue`
- Push de cada rama
- Limpieza de worktrees
- Logging en `logs/`
**Flags disponibles:**
- `--sort` → analizar issues y generar PARALLEL_EXECUTION_ORDER.md
- `--dry-run` → mostrar plan y worktrees que se crearían
- `--group N` → ejecutar solo grupo N
- `--sequential` → ejecutar sin paralelismo
- `--timeout N` → timeout en minutos por issue (default: 30)
- `--cleanup` → solo limpiar worktrees existentes
- `--plan file` → usar plan alternativo
### 4. Mostrar resumen
### 3. Mostrar resumen
Después de ejecutar, mostrar:
- Resultados por issue (éxito/fallo)
- Ruta a logs y summary
- Estado de worktrees
### 4. Limpiar plan si exitoso
Si todos los issues completaron exitosamente, eliminar `PARALLEL_EXECUTION_ORDER.md`.
## Arquitectura
```
Ejecución completada
Logs:
- logs/parallel-execution-*.log
- logs/consolidated-summary.txt
Worktrees restantes: (ninguno)
```
### 5. Eliminar plan
Si exitoso, eliminar `PARALLEL_EXECUTION_ORDER.md`.
## Arquitectura Go
```
cmd/parallel-executor/
├── main.go # CLI
├── parser.go # Parse plan
├── worktree.go # Git worktrees
├── executor.go # Ejecutar claude
├── logger.go # Logging
└── orchestrator.go # Goroutines
utils/parallel-executor/
├── main.go # CLI + orquestación
├── core/
│ ├── parser.go # Parseo del plan markdown (puro)
│ ├── planner.go # Topological sort + conflictos (puro)
│ ├── parser_test.go
│ └── planner_test.go
├── shell/
│ ├── worktree.go # Git worktree CRUD
│ ├── executor.go # Ejecutar claude en worktree
│ └── logger.go # Logging a disco
├── go.mod, go.work # DevFactory como dependencia
└── Makefile
```
## Convenciones
- Logs persistentes por ejecución
- Timeout 30 min por issue
- Limpieza automática de worktrees
- Plan se elimina al completar
- Usa DevFactory: `Result[T]`, `MapSlice`, `FilterSlice`, `Reduce`
- Patrón pure core / impure shell
- Logs persistentes en `logs/`
- Timeout 30 min por issue (configurable)
- Limpieza automática de worktrees al terminar
- Plan se auto-genera si no existe (`--sort` implícito)
## Reglas
- SIEMPRE generar plan si no existe
- Solo advertir si hay cambios (no bloquear)
- SIEMPRE verificar que el binario existe antes de ejecutar
- SIEMPRE mostrar dry-run antes de una ejecución real si el usuario no especificó flags
- SIEMPRE limpiar worktrees al terminar
- Si no hay plan, generarlo automáticamente
+62
View File
@@ -0,0 +1,62 @@
---
name: init-frontend
description: Inicializa proyecto frontend (React/Vite) o desktop (Wails) con Frontend_Library
disable-model-invocation: true
user-invocable: true
allowed-tools: Bash, Read, Write, Edit
---
# init-frontend
Inicializa un proyecto frontend (webapp React/Vite) o desktop (Wails + Go + React). Coherente con Frontend_Library y el stack del frontend-lib/build-wails agents.
## Sintaxis
```bash
/init-frontend [nombre] [--wails] [--path /ruta/destino]
```
- `nombre`: nombre del proyecto (kebab-case). Si no se da, se pregunta.
- `--wails`: modo desktop con Wails (Go backend + React frontend). Sin flag = webapp pura.
- `--path`: directorio destino. Default: directorio actual.
## Flujo
### 1. Ejecutar script de setup
```bash
bash "${CLAUDE_SKILL_DIR}/setup-frontend.sh" [nombre] [--wails] [path]
```
### 2. Si el script reporta STATUS: CONFIGURED
Informar al usuario que el proyecto ya existe.
### 3. Si el script reporta STATUS: READY
Mostrar resumen según modo:
**Webapp:**
- `pnpm dev` para desarrollo
- `pnpm build` para producción
- Frontend_Library linkeada via pnpm
**Wails:**
- `make dev` para desarrollo con hot reload
- `make build` para compilar
- Frontend_Library + DevFactory integrados
- Bindings Go→TS auto-generados
### 4. Si el script reporta STATUS: ERROR
Mostrar el error y sugerir corrección.
## Convenciones
- pnpm exclusivamente (no npm ni yarn)
- React 19 + TypeScript + Vite + Tailwind CSS 4
- @anthropic/frontend-lib via pnpm link
- Temas OKLCH con semantic tokens
- Phosphor Icons
- Vite dedupe obligatorio para react/react-dom
- En modo Wails: go.work con DevFactory, patrón pure core / impure shell
+438
View File
@@ -0,0 +1,438 @@
#!/usr/bin/env bash
set -euo pipefail
# =============================================================================
# setup-frontend.sh — Inicializa proyecto React/Vite o Wails desktop
# Coherente con Frontend_Library + DevFactory + build-wails agent
# =============================================================================
# --- Colores ---
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
BLUE='\033[0;34m'
CYAN='\033[0;36m'
NC='\033[0m'
log_info() { echo -e "${BLUE}[INFO]${NC} $1"; }
log_ok() { echo -e "${GREEN}[OK]${NC} $1"; }
log_warn() { echo -e "${YELLOW}[WARN]${NC} $1"; }
log_error() { echo -e "${RED}[ERROR]${NC} $1"; }
log_step() { echo -e "${CYAN}[STEP]${NC} $1"; }
# --- Parámetros ---
PROJECT_NAME=""
WAILS_MODE=false
TARGET_PATH="."
while [[ $# -gt 0 ]]; do
case "$1" in
--wails) WAILS_MODE=true; shift ;;
--path) TARGET_PATH="$2"; shift 2 ;;
-*) log_error "Flag desconocido: $1"; echo "STATUS: ERROR"; exit 1 ;;
*) PROJECT_NAME="$1"; shift ;;
esac
done
# --- Rutas de librerías ---
FRONTEND_LIB="$HOME/.local_agentes/frontend/frontend"
DEVFACTORY_PATH="$HOME/.local_agentes/backend"
TEMPLATES_DIR="$HOME/.local_agentes/frontend/templates/base"
WAILS_TEMPLATES="$HOME/.claude/agents/build-wails/templates"
# --- Validar nombre ---
if [[ -z "$PROJECT_NAME" ]]; then
log_error "Uso: setup-frontend.sh <nombre> [--wails] [--path /ruta]"
echo "STATUS: ERROR"
exit 1
fi
# Normalizar
PROJECT_NAME=$(echo "$PROJECT_NAME" | tr '[:upper:]' '[:lower:]' | sed 's/[^a-z0-9-]/-/g' | sed 's/--*/-/g' | sed 's/^-\|-$//g')
PROJECT_DIR="$TARGET_PATH/$PROJECT_NAME"
# --- Check estado existente ---
if [[ -f "$PROJECT_DIR/package.json" ]] || [[ -f "$PROJECT_DIR/wails.json" ]]; then
log_warn "El proyecto $PROJECT_NAME ya existe en $PROJECT_DIR"
echo "STATUS: CONFIGURED"
exit 0
fi
# --- Verificar dependencias ---
log_step "Verificando dependencias..."
if ! command -v pnpm &>/dev/null; then
log_error "pnpm no está instalado. Instala con: npm install -g pnpm"
echo "STATUS: ERROR"
exit 1
fi
log_ok "pnpm $(pnpm --version) encontrado"
if ! command -v node &>/dev/null; then
log_error "Node.js no encontrado"
echo "STATUS: ERROR"
exit 1
fi
log_ok "Node $(node --version) encontrado"
if [[ "$WAILS_MODE" == true ]]; then
if ! command -v wails &>/dev/null; then
log_error "Wails no está instalado. Instala con: go install github.com/wailsapp/wails/v2/cmd/wails@latest"
echo "STATUS: ERROR"
exit 1
fi
log_ok "Wails $(wails version 2>/dev/null | head -1 || echo 'v2.x') encontrado"
if ! command -v go &>/dev/null; then
log_error "Go no encontrado (requerido para Wails)"
echo "STATUS: ERROR"
exit 1
fi
log_ok "Go $(go version | grep -oP '\d+\.\d+' | head -1) encontrado"
fi
if [[ ! -d "$FRONTEND_LIB" ]]; then
log_warn "Frontend_Library no encontrada en $FRONTEND_LIB — se creará sin link"
HAS_FRONTEND_LIB=false
else
log_ok "Frontend_Library encontrada"
HAS_FRONTEND_LIB=true
fi
# ============================================================================
# MODO WAILS — Desktop app (Go + React)
# ============================================================================
if [[ "$WAILS_MODE" == true ]]; then
log_step "Creando proyecto Wails '$PROJECT_NAME'..."
# Usar wails init con template react-ts
wails init -n "$PROJECT_NAME" -t react-ts -d "$TARGET_PATH" 2>/dev/null
cd "$PROJECT_DIR"
# --- go.work con DevFactory ---
if [[ -d "$DEVFACTORY_PATH" ]]; then
log_step "Configurando go.work con DevFactory..."
cat > go.work << EOF
go 1.22
use (
.
$DEVFACTORY_PATH
)
EOF
log_ok "DevFactory enlazado via go.work"
fi
# --- Configurar frontend con pnpm ---
log_step "Configurando frontend con pnpm..."
cd frontend
# Reemplazar npm por pnpm en wails.json
cd ..
if [[ -f "wails.json" ]]; then
sed -i 's/"npm install"/"pnpm install"/g' wails.json
sed -i 's/"npm run dev"/"pnpm dev"/g' wails.json
sed -i 's/"npm run build"/"pnpm build"/g' wails.json
fi
cd frontend
# Instalar con pnpm
rm -f package-lock.json 2>/dev/null || true
pnpm install
# --- Linkear Frontend_Library ---
if [[ "$HAS_FRONTEND_LIB" == true ]]; then
log_step "Linkeando Frontend_Library..."
pnpm add "@anthropic/frontend-lib@link:$FRONTEND_LIB"
log_ok "@anthropic/frontend-lib linkeada"
fi
# --- Instalar Tailwind CSS 4 ---
log_step "Instalando Tailwind CSS 4..."
pnpm add -D tailwindcss @tailwindcss/vite
# --- Configurar vite.config.ts con dedupe y tailwind ---
log_step "Configurando Vite (dedupe + tailwind)..."
cat > vite.config.ts << 'VEOF'
import { defineConfig } from 'vite'
import react from '@vitejs/plugin-react'
import tailwindcss from '@tailwindcss/vite'
import { resolve } from 'path'
export default defineConfig({
plugins: [react(), tailwindcss()],
resolve: {
alias: {
'@': resolve(__dirname, './src'),
'@wails': resolve(__dirname, './wailsjs'),
},
dedupe: ['react', 'react-dom'],
},
})
VEOF
# --- CSS base con Tailwind ---
cat > src/style.css << 'CSSEOF'
@import "tailwindcss";
CSSEOF
# --- Instalar Phosphor Icons ---
pnpm add @phosphor-icons/react
cd ..
# --- Makefile (basado en build-wails agent) ---
log_step "Generando Makefile..."
cat > Makefile << 'MKEOF'
.PHONY: dev dev-debug build build-prod build-linux build-windows build-all clean generate doctor
APP_NAME := $(shell basename $(CURDIR))
## dev: Desarrollo con hot reload
dev:
wails dev
## dev-debug: Desarrollo con DevTools
dev-debug:
wails dev -devtools
## build: Build para plataforma actual
build:
wails build
## build-prod: Build optimizado para producción
build-prod:
wails build -clean -trimpath -ldflags="-s -w"
## build-linux: Build para Linux AMD64
build-linux:
wails build -platform linux/amd64
## build-windows: Cross-compile para Windows
build-windows:
wails build -platform windows/amd64
## build-all: Linux + Windows
build-all: build-linux build-windows
## generate: Regenerar bindings TypeScript
generate:
wails generate module
## clean: Limpiar artefactos
clean:
rm -rf build/bin frontend/dist
## doctor: Verificar instalación
doctor:
wails doctor
## help: Muestra esta ayuda
help:
@grep -E '^## ' Makefile | sed 's/## //' | column -t -s ':'
MKEOF
# --- .gitignore ---
cat >> .gitignore << 'EOF'
node_modules/
frontend/dist/
build/bin/
*.exe
EOF
# --- Resumen Wails ---
echo ""
log_ok "Proyecto Wails '$PROJECT_NAME' creado en $PROJECT_DIR"
echo ""
echo -e "${CYAN}Estructura:${NC}"
echo " $PROJECT_NAME/"
echo " ├── main.go, app.go — Backend Go"
echo " ├── go.work — Enlace a DevFactory"
echo " ├── frontend/ — React + TypeScript + Vite"
echo " │ ├── src/ — Componentes React"
echo " │ └── wailsjs/ — Bindings auto-generados"
echo " ├── Makefile — dev, build, build-all"
echo " └── wails.json — Configuración Wails"
echo ""
echo -e "${CYAN}Comandos:${NC}"
echo " make dev — Desarrollo con hot reload"
echo " make build — Compilar app desktop"
echo " make build-all — Linux + Windows"
echo " make generate — Regenerar bindings TS"
echo ""
if [[ "$HAS_FRONTEND_LIB" == true ]]; then
echo -e "${CYAN}Frontend_Library:${NC}"
echo " import { Button, Card } from '@anthropic/frontend-lib'"
echo " import { useTheme } from '@anthropic/frontend-lib/hooks'"
echo ""
fi
echo "STATUS: READY"
exit 0
fi
# ============================================================================
# MODO WEBAPP — React + Vite (sin Wails)
# ============================================================================
log_step "Creando proyecto webapp '$PROJECT_NAME'..."
mkdir -p "$PROJECT_DIR"
cd "$PROJECT_DIR"
# --- Usar template de Frontend_Library si existe ---
if [[ -d "$TEMPLATES_DIR" ]]; then
log_step "Usando template de Frontend_Library..."
cp -r "$TEMPLATES_DIR"/* .
cp -r "$TEMPLATES_DIR"/.[!.]* . 2>/dev/null || true
else
log_step "Creando desde cero con Vite..."
# package.json
cat > package.json << PKGEOF
{
"name": "$PROJECT_NAME",
"private": true,
"version": "0.1.0",
"type": "module",
"scripts": {
"dev": "vite",
"build": "tsc -b && vite build",
"preview": "vite preview"
}
}
PKGEOF
# Instalar dependencias base
pnpm add react react-dom
pnpm add -D typescript @types/react @types/react-dom
pnpm add -D vite @vitejs/plugin-react
pnpm add -D tailwindcss @tailwindcss/vite
# tsconfig.json
cat > tsconfig.json << 'TSEOF'
{
"compilerOptions": {
"target": "ES2023",
"lib": ["ES2023", "DOM", "DOM.Iterable"],
"module": "ESNext",
"moduleResolution": "bundler",
"jsx": "react-jsx",
"strict": true,
"noUnusedLocals": true,
"noUnusedParameters": true,
"paths": { "@/*": ["./src/*"] },
"skipLibCheck": true
},
"include": ["src"]
}
TSEOF
# vite.config.ts
cat > vite.config.ts << 'VEOF'
import { defineConfig } from 'vite'
import react from '@vitejs/plugin-react'
import tailwindcss from '@tailwindcss/vite'
import { resolve } from 'path'
export default defineConfig({
plugins: [react(), tailwindcss()],
resolve: {
alias: { '@': resolve(__dirname, './src') },
dedupe: ['react', 'react-dom'],
},
})
VEOF
# index.html
cat > index.html << HTMLEOF
<!doctype html>
<html lang="en">
<head>
<meta charset="UTF-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<title>$PROJECT_NAME</title>
</head>
<body>
<div id="root"></div>
<script type="module" src="/src/main.tsx"></script>
</body>
</html>
HTMLEOF
# src/
mkdir -p src
cat > src/main.tsx << 'TSXEOF'
import { StrictMode } from 'react'
import { createRoot } from 'react-dom/client'
import App from './App'
import './app.css'
createRoot(document.getElementById('root')!).render(
<StrictMode>
<App />
</StrictMode>,
)
TSXEOF
cat > src/App.tsx << 'TSXEOF'
function App() {
return (
<div className="min-h-screen bg-surface text-foreground flex items-center justify-center">
<h1 className="text-3xl font-bold">Ready</h1>
</div>
)
}
export default App
TSXEOF
cat > src/app.css << 'CSSEOF'
@import "tailwindcss";
CSSEOF
fi
# --- Linkear Frontend_Library ---
if [[ "$HAS_FRONTEND_LIB" == true ]]; then
log_step "Linkeando Frontend_Library..."
pnpm add "@anthropic/frontend-lib@link:$FRONTEND_LIB"
log_ok "@anthropic/frontend-lib linkeada"
fi
# --- Phosphor Icons ---
pnpm add @phosphor-icons/react
# --- .gitignore ---
cat > .gitignore << 'EOF'
node_modules/
dist/
.vite/
*.local
EOF
# --- Resumen Webapp ---
echo ""
log_ok "Proyecto webapp '$PROJECT_NAME' creado en $PROJECT_DIR"
echo ""
echo -e "${CYAN}Estructura:${NC}"
echo " $PROJECT_NAME/"
echo " ├── src/"
echo " │ ├── App.tsx — Componente principal"
echo " │ ├── main.tsx — Entry point"
echo " │ └── app.css — Tailwind CSS"
echo " ├── vite.config.ts — Vite + Tailwind + dedupe"
echo " ├── tsconfig.json — TypeScript strict"
echo " └── package.json — pnpm"
echo ""
echo -e "${CYAN}Comandos:${NC}"
echo " pnpm dev — Servidor de desarrollo"
echo " pnpm build — Build de producción"
echo " pnpm preview — Preview del build"
echo ""
if [[ "$HAS_FRONTEND_LIB" == true ]]; then
echo -e "${CYAN}Frontend_Library:${NC}"
echo " import { Button, Card } from '@anthropic/frontend-lib'"
echo " import { useTheme } from '@anthropic/frontend-lib/hooks'"
echo ""
fi
echo "STATUS: READY"
+53
View File
@@ -0,0 +1,53 @@
---
name: init-go-module
description: Inicializa un módulo Go funcional con bindings Python (CGO c-shared + ctypes)
disable-model-invocation: true
user-invocable: true
allowed-tools: Bash, Read, Write, Edit
---
# init-go-module
Inicializa un módulo Go con arquitectura funcional (pure core / impure shell) y bindings Python automáticos via CGO c-shared + ctypes. Coherente con DevFactory y el stack del backend-lib agent.
## Sintaxis
```bash
/init-go-module [nombre] [--path /ruta/destino]
```
- `nombre`: nombre del módulo (kebab-case). Si no se da, se pregunta.
- `--path`: directorio destino. Default: directorio actual.
## Flujo
### 1. Ejecutar script de setup
```bash
bash "${CLAUDE_SKILL_DIR}/setup-go-module.sh" [nombre] [path]
```
### 2. Si el script reporta STATUS: CONFIGURED
Informar al usuario que el módulo ya está configurado.
### 3. Si el script reporta STATUS: READY
Mostrar resumen:
- Estructura creada
- Cómo compilar: `make build`
- Cómo generar bindings Python: `make python`
- Cómo testear: `make test`
- Cómo usar desde Python: `from bindings.modulo import *`
### 4. Si el script reporta STATUS: ERROR
Mostrar el error y sugerir corrección.
## Convenciones
- Usa DevFactory como dependencia via `go.work` (igual que build-wails)
- Patrón pure core / impure shell de DevFactory
- `Result[T]` y `Option[T]` del core de DevFactory
- Funciones exportadas a Python son thin wrappers en `export/`
- El wrapper Python se auto-genera desde los `//export` comments
+466
View File
@@ -0,0 +1,466 @@
#!/usr/bin/env bash
set -euo pipefail
# =============================================================================
# setup-go-module.sh — Inicializa módulo Go funcional con bindings Python
# Coherente con DevFactory (pure core / impure shell) + CGO c-shared + ctypes
# =============================================================================
# --- Colores ---
RED='\033[0;31m'
GREEN='\033[0;32m'
YELLOW='\033[1;33m'
BLUE='\033[0;34m'
CYAN='\033[0;36m'
NC='\033[0m'
log_info() { echo -e "${BLUE}[INFO]${NC} $1"; }
log_ok() { echo -e "${GREEN}[OK]${NC} $1"; }
log_warn() { echo -e "${YELLOW}[WARN]${NC} $1"; }
log_error() { echo -e "${RED}[ERROR]${NC} $1"; }
log_step() { echo -e "${CYAN}[STEP]${NC} $1"; }
# --- Parámetros ---
MODULE_NAME="${1:-}"
TARGET_PATH="${2:-.}"
DEVFACTORY_PATH="$HOME/.local_agentes/backend"
DEVFACTORY_MODULE="github.com/lucasdataproyects/devfactory"
# --- Validar nombre ---
if [[ -z "$MODULE_NAME" ]]; then
log_error "Uso: setup-go-module.sh <nombre> [path]"
echo "STATUS: ERROR"
exit 1
fi
# Normalizar nombre a kebab-case
MODULE_NAME=$(echo "$MODULE_NAME" | tr '[:upper:]' '[:lower:]' | sed 's/[^a-z0-9-]/-/g' | sed 's/--*/-/g' | sed 's/^-\|-$//g')
PROJECT_DIR="$TARGET_PATH/$MODULE_NAME"
# --- Check estado existente ---
if [[ -f "$PROJECT_DIR/go.mod" ]]; then
log_warn "El módulo $MODULE_NAME ya existe en $PROJECT_DIR"
if [[ -f "$PROJECT_DIR/export/exports.go" ]]; then
log_ok "Bindings Python ya configurados"
else
log_warn "Falta directorio export/ — ejecuta de nuevo para completar"
fi
echo "STATUS: CONFIGURED"
exit 0
fi
# --- Verificar dependencias ---
log_step "Verificando dependencias..."
if ! command -v go &>/dev/null; then
log_error "Go no está instalado. Instala Go 1.22+"
echo "STATUS: ERROR"
exit 1
fi
GO_VERSION=$(go version | grep -oP '\d+\.\d+' | head -1)
log_ok "Go $GO_VERSION encontrado"
if ! command -v python3 &>/dev/null; then
log_warn "Python3 no encontrado — los bindings se generarán pero no se podrán testear"
fi
if [[ ! -d "$DEVFACTORY_PATH" ]]; then
log_warn "DevFactory no encontrado en $DEVFACTORY_PATH — se creará go.mod sin go.work"
fi
# --- Crear estructura ---
log_step "Creando estructura del módulo '$MODULE_NAME'..."
mkdir -p "$PROJECT_DIR"/{core,shell,export,python/bindings,cmd,internal}
# --- go.mod ---
log_step "Generando go.mod..."
cat > "$PROJECT_DIR/go.mod" << EOF
module github.com/lucasdataproyects/$MODULE_NAME
go 1.22
require $DEVFACTORY_MODULE v0.0.0
EOF
# --- go.work (si DevFactory existe localmente) ---
if [[ -d "$DEVFACTORY_PATH" ]]; then
log_step "Generando go.work con DevFactory local..."
cat > "$PROJECT_DIR/go.work" << EOF
go 1.22
use (
.
$DEVFACTORY_PATH
)
EOF
log_ok "go.work enlazado a DevFactory"
fi
# --- core/transform.go — Funciones puras de ejemplo ---
log_step "Generando core/ (funciones puras)..."
cat > "$PROJECT_DIR/core/transform.go" << 'GOEOF'
// Package core contiene funciones puras sin side effects.
// Todas las funciones son deterministas y composables.
package core
import (
"strings"
"errors"
df "github.com/lucasdataproyects/devfactory/core"
)
// ToUpper transforma texto a mayúsculas (función pura).
func ToUpper(s string) string {
return strings.ToUpper(s)
}
// ProcessItems aplica una transformación a cada elemento usando MapSlice de DevFactory.
func ProcessItems(items []string, transform func(string) string) []string {
return df.MapSlice(items, transform)
}
// FilterNonEmpty filtra elementos vacíos usando FilterSlice de DevFactory.
func FilterNonEmpty(items []string) []string {
return df.FilterSlice(items, func(s string) bool {
return len(strings.TrimSpace(s)) > 0
})
}
// SafeDivide retorna Result[float64] para evitar panic en división por cero.
func SafeDivide(a, b float64) df.Result[float64] {
if b == 0 {
return df.Err[float64](errors.New("division by zero"))
}
return df.Ok(a / b)
}
GOEOF
# --- core/types.go — Tipos exportables a Python ---
cat > "$PROJECT_DIR/core/types.go" << 'GOEOF'
package core
// DataPoint representa un punto de datos exportable a Python.
// Los campos usan tipos C-compatible para facilitar el binding.
type DataPoint struct {
Label string
Value float64
}
// Summary es el resultado de un procesamiento, exportable a Python.
type Summary struct {
Count int
Total float64
Items []string
}
GOEOF
# --- shell/io.go — Operaciones I/O con Result[T] ---
log_step "Generando shell/ (operaciones I/O)..."
cat > "$PROJECT_DIR/shell/io.go" << 'GOEOF'
// Package shell contiene operaciones con side effects, wrapeadas en Result[T].
package shell
import (
df "github.com/lucasdataproyects/devfactory/core"
"github.com/lucasdataproyects/devfactory/shell"
)
// ReadDataFile lee un archivo y retorna su contenido como Result.
func ReadDataFile(path string) df.Result[string] {
return shell.ReadString(path)
}
// WriteResult escribe un resultado a archivo.
func WriteResult(path string, content string) df.Result[struct{}] {
return shell.WriteString(path, content)
}
GOEOF
# --- export/exports.go — Funciones exportadas via CGO ---
log_step "Generando export/ (bindings CGO)..."
cat > "$PROJECT_DIR/export/exports.go" << GOEOF
// Package main exporta funciones Go como C shared library.
// Cada función con //export se expone como símbolo C callable desde Python.
package main
import "C"
import (
"encoding/json"
"unsafe"
"github.com/lucasdataproyects/$MODULE_NAME/core"
)
//export GoToUpper
func GoToUpper(input *C.char) *C.char {
result := core.ToUpper(C.GoString(input))
return C.CString(result)
}
//export GoProcessItems
func GoProcessItems(jsonInput *C.char) *C.char {
var items []string
if err := json.Unmarshal([]byte(C.GoString(jsonInput)), &items); err != nil {
return C.CString("[]")
}
result := core.ProcessItems(items, core.ToUpper)
out, _ := json.Marshal(result)
return C.CString(string(out))
}
//export GoFilterNonEmpty
func GoFilterNonEmpty(jsonInput *C.char) *C.char {
var items []string
if err := json.Unmarshal([]byte(C.GoString(jsonInput)), &items); err != nil {
return C.CString("[]")
}
result := core.FilterNonEmpty(items)
out, _ := json.Marshal(result)
return C.CString(string(out))
}
//export GoSafeDivide
func GoSafeDivide(a, b C.double) *C.char {
result := core.SafeDivide(float64(a), float64(b))
if result.IsErr() {
return C.CString(`{"error":"` + result.Error().Error() + `"}`)
}
out, _ := json.Marshal(map[string]float64{"value": result.Unwrap()})
return C.CString(string(out))
}
//export GoFree
func GoFree(ptr *C.char) {
C.free(unsafe.Pointer(ptr))
}
func main() {}
GOEOF
# --- python/bindings/__init__.py — Wrapper ctypes auto-generado ---
log_step "Generando python/bindings/ (ctypes wrapper)..."
# Nombre de la shared library según OS
SO_NAME="lib${MODULE_NAME}.so"
cat > "$PROJECT_DIR/python/bindings/__init__.py" << PYEOF
"""
Auto-generated Python bindings for $MODULE_NAME.
Uses ctypes to call Go functions compiled as C shared library.
Usage:
from bindings import to_upper, process_items, filter_non_empty, safe_divide
"""
import ctypes
import json
import os
from pathlib import Path
# Localizar la shared library
_LIB_DIR = Path(__file__).parent.parent.parent / "build"
_LIB_NAME = "$SO_NAME"
_LIB_PATH = _LIB_DIR / _LIB_NAME
if not _LIB_PATH.exists():
raise FileNotFoundError(
f"Shared library not found at {_LIB_PATH}. "
f"Run 'make build' in the project root first."
)
_lib = ctypes.CDLL(str(_LIB_PATH))
# --- Configurar tipos de retorno ---
_lib.GoToUpper.argtypes = [ctypes.c_char_p]
_lib.GoToUpper.restype = ctypes.c_char_p
_lib.GoProcessItems.argtypes = [ctypes.c_char_p]
_lib.GoProcessItems.restype = ctypes.c_char_p
_lib.GoFilterNonEmpty.argtypes = [ctypes.c_char_p]
_lib.GoFilterNonEmpty.restype = ctypes.c_char_p
_lib.GoSafeDivide.argtypes = [ctypes.c_double, ctypes.c_double]
_lib.GoSafeDivide.restype = ctypes.c_char_p
_lib.GoFree.argtypes = [ctypes.c_char_p]
_lib.GoFree.restype = None
def to_upper(text: str) -> str:
"""Convert text to uppercase using Go core."""
result = _lib.GoToUpper(text.encode("utf-8"))
return result.decode("utf-8")
def process_items(items: list[str]) -> list[str]:
"""Process items through Go pipeline (ToUpper transformation)."""
input_json = json.dumps(items).encode("utf-8")
result = _lib.GoProcessItems(input_json)
return json.loads(result.decode("utf-8"))
def filter_non_empty(items: list[str]) -> list[str]:
"""Filter empty strings using Go core."""
input_json = json.dumps(items).encode("utf-8")
result = _lib.GoFilterNonEmpty(input_json)
return json.loads(result.decode("utf-8"))
def safe_divide(a: float, b: float) -> float:
"""Safe division using Go Result type. Raises ValueError on division by zero."""
result = _lib.GoSafeDivide(ctypes.c_double(a), ctypes.c_double(b))
data = json.loads(result.decode("utf-8"))
if "error" in data:
raise ValueError(data["error"])
return data["value"]
PYEOF
# --- python/example.py ---
cat > "$PROJECT_DIR/python/example.py" << PYEOF
"""Example usage of $MODULE_NAME Go bindings from Python."""
from bindings import to_upper, process_items, filter_non_empty, safe_divide
# String transformation
print(to_upper("hello from go")) # HELLO FROM GO
# Batch processing via Go's MapSlice
items = ["hello", "world", "from", "go"]
print(process_items(items)) # ["HELLO", "WORLD", "FROM", "GO"]
# Filtering via Go's FilterSlice
mixed = ["hello", "", "world", " ", "go"]
print(filter_non_empty(mixed)) # ["hello", "world", "go"]
# Safe division with Result[T] error handling
print(safe_divide(10.0, 3.0)) # 3.333...
try:
safe_divide(10.0, 0.0)
except ValueError as e:
print(f"Caught: {e}") # Caught: division by zero
PYEOF
# --- core/transform_test.go ---
log_step "Generando tests..."
cat > "$PROJECT_DIR/core/transform_test.go" << 'GOEOF'
package core
import (
"testing"
)
func TestToUpper(t *testing.T) {
if got := ToUpper("hello"); got != "HELLO" {
t.Errorf("ToUpper(\"hello\") = %q, want %q", got, "HELLO")
}
}
func TestFilterNonEmpty(t *testing.T) {
items := []string{"hello", "", "world", " ", "go"}
result := FilterNonEmpty(items)
if len(result) != 3 {
t.Errorf("FilterNonEmpty got %d items, want 3", len(result))
}
}
func TestSafeDivide(t *testing.T) {
ok := SafeDivide(10, 2)
if ok.IsErr() {
t.Error("SafeDivide(10, 2) should not error")
}
if ok.Unwrap() != 5.0 {
t.Errorf("SafeDivide(10, 2) = %f, want 5.0", ok.Unwrap())
}
err := SafeDivide(10, 0)
if !err.IsErr() {
t.Error("SafeDivide(10, 0) should error")
}
}
GOEOF
# --- Makefile ---
log_step "Generando Makefile..."
cat > "$PROJECT_DIR/Makefile" << MKEOF
.PHONY: build test clean python dev
MODULE_NAME := $MODULE_NAME
BUILD_DIR := build
SO_NAME := $SO_NAME
## build: Compila la shared library (.so) para Python
build:
@mkdir -p \$(BUILD_DIR)
cd export && CGO_ENABLED=1 go build -buildmode=c-shared -o ../\$(BUILD_DIR)/\$(SO_NAME) .
@echo "✓ Built \$(BUILD_DIR)/\$(SO_NAME)"
## test: Ejecuta tests de Go
test:
go test ./core/... ./shell/... -v
## python: Compila y ejecuta ejemplo Python
python: build
cd python && python3 example.py
## clean: Limpia artefactos
clean:
rm -rf \$(BUILD_DIR)
find . -name "__pycache__" -type d -exec rm -rf {} + 2>/dev/null || true
## dev: Tests + build en un paso
dev: test build
@echo "✓ Ready — run 'make python' to test bindings"
## tidy: go mod tidy
tidy:
go mod tidy
## help: Muestra esta ayuda
help:
@grep -E '^## ' Makefile | sed 's/## //' | column -t -s ':'
MKEOF
# --- .gitignore ---
cat > "$PROJECT_DIR/.gitignore" << 'EOF'
build/
*.so
*.h
*.dylib
*.dll
__pycache__/
*.pyc
.pytest_cache/
EOF
# --- go mod tidy ---
log_step "Ejecutando go mod tidy..."
cd "$PROJECT_DIR"
if [[ -f "go.work" ]]; then
go mod tidy 2>/dev/null || log_warn "go mod tidy falló — revisa el go.work"
else
go mod tidy 2>/dev/null || log_warn "go mod tidy falló — DevFactory no está disponible"
fi
# --- Resumen ---
echo ""
log_ok "Módulo '$MODULE_NAME' creado en $PROJECT_DIR"
echo ""
echo -e "${CYAN}Estructura:${NC}"
echo " $MODULE_NAME/"
echo " ├── core/ — Funciones puras (sin side effects)"
echo " ├── shell/ — Operaciones I/O con Result[T]"
echo " ├── export/ — Funciones exportadas via CGO"
echo " ├── python/bindings — Wrapper ctypes auto-generado"
echo " ├── Makefile — build, test, python, clean"
echo " └── go.work — Enlace a DevFactory"
echo ""
echo -e "${CYAN}Comandos:${NC}"
echo " make test — Ejecutar tests Go"
echo " make build — Compilar shared library"
echo " make python — Testear bindings Python"
echo " make dev — Test + build en un paso"
echo ""
echo "STATUS: READY"
+209
View File
@@ -0,0 +1,209 @@
# Claude Code — Skills, Agents & Tools
Sistema de automatizacion para desarrollo de software usando Claude Code. Incluye skills (comandos invocables), agentes especializados, y herramientas Go para orquestar trabajo paralelo.
## Estructura del repo
```
repo_Claude/
├── bin/ # Binarios compilados
│ └── parallel-executor # Orquestador de ejecucion paralela
├── utils/
│ └── parallel-executor/ # Codigo fuente Go del orquestador
│ ├── core/ # Funciones puras (parser, planner)
│ └── shell/ # I/O (worktrees, executor, logger)
├── dev/
│ └── issues/ # Sistema local de issues
└── install.sh # Instalador
```
---
## Skills
Skills son comandos invocables desde Claude Code con `/nombre`. Viven en `~/.claude/skills/`.
### Configuracion y setup
| Skill | Descripcion | Uso |
|-------|-------------|-----|
| `/primer` | Genera CLAUDE.md personalizado analizando el repo | `/primer` |
| `/init-jupyter` | Inicializa proyecto Jupyter + MCP (bash script idempotente) | `/init-jupyter [ruta]` |
| `/init-go-module` | Crea modulo Go funcional con bindings Python (CGO + ctypes) | `/init-go-module nombre` |
| `/init-frontend` | Crea proyecto React/Vite o Wails desktop | `/init-frontend nombre [--wails]` |
| `/nochanges` | Modo read-only para explorar sin modificar | `/nochanges` |
| `/create-skill` | Crea un skill nuevo | `/create-skill nombre` |
| `/create-agent` | Crea un agente especializado | `/create-agent` |
### Git
| Skill | Descripcion | Uso |
|-------|-------------|-----|
| `/git-branch` | Crea ramas `issue/*` o `quick/*` desde master | `/git-branch issue 0013 slug` |
| `/git-push` | Commits atomicos por tipo, merge --no-ff, push, limpieza | `/git-push` |
| `/git-recovery` | Recupera repo de estados inconsistentes | `/git-recovery [--aggressive]` |
### Workspace (Gitea + SQLite)
| Skill | Descripcion | Uso |
|-------|-------------|-----|
| `/create-repo` | Crea workspace en Gitea con rollback | `/create-repo` |
| `/import-repo` | Importa repo existente a Gitea (mirror) | `/import-repo` |
| `/sync-repos` | Sincroniza workspaces locales con Gitea | `/sync-repos [--dry-run]` |
| `/list-repos` | Lista workspaces desde SQLite | `/list-repos [--filter x]` |
| `/cleanup-worktrees` | Limpia worktrees post-merge | `/cleanup-worktrees [--all]` |
### Issues
| Skill | Descripcion | Uso |
|-------|-------------|-----|
| `/create-issue` | Crea issue con template, sub-issues si es grande | `/create-issue` |
| `/fix-issue` | E2E: lee issue, branch, implementa, tests, merge | `/fix-issue 0013` |
| `/auto-fix` | Igual que fix-issue pero sin confirmacion | `/auto-fix 0013` |
| `/auto-create` | Igual que create-issue sin confirmacion | `/auto-create` |
| `/quick-issue` | Issue minimal desde texto (para TUI) | `/quick-issue --text "..."` |
| `/issues-status` | Dashboard con filtros por workspace/estado/tag | `/issues-status` |
### Ejecucion paralela
| Skill | Descripcion | Uso |
|-------|-------------|-----|
| `/sort-issues` | Analiza dependencias, topological sort | `/sort-issues` |
| `/parallel-issues` | Genera plan de ejecucion paralela | `/parallel-issues` |
| `/execute-parallel` | Crea worktrees y ejecuta issues en paralelo | `/execute-parallel [--dry-run]` |
---
## Agentes
Agentes especializados que Claude Code puede invocar automaticamente. Cada uno tiene su propio modelo, herramientas y MCP servers. Viven en `~/.claude/agents/`.
### Librerias de desarrollo
| Agente | Modelo | Descripcion | Ubicacion |
|--------|--------|-------------|-----------|
| **backend-lib** | sonnet | Gestiona DevFactory — libreria Go funcional con Result[T], Option[T], HTTP, DB, finance | `~/.local_agentes/backend` |
| **frontend-lib** | sonnet | Gestiona Frontend_Library — 50+ componentes React/TS, temas OKLCH, shadcn/ui, charts, DSP | `~/.local_agentes/frontend` |
### Build y deploy
| Agente | Modelo | Descripcion |
|--------|--------|-------------|
| **build-wails** | sonnet | Apps desktop Wails v2 (Go + React). Cross-compile Linux/Windows/macOS. Integra ambas librerias |
| **docker** | sonnet | Dockerfiles multi-stage, docker-compose, push a registries, deploy via SSH |
### Datos e infraestructura
| Agente | Modelo | Descripcion |
|--------|--------|-------------|
| **db-reader** | sonnet | Bases de datos SQLite y DuckDB. Consultas, imports CSV/Parquet/JSON, analisis OLAP |
| **gitea** | sonnet | Gestion de Gitea: repos, issues, PRs, branches, archivos. MCP integrado |
### Automatizacion
| Agente | Modelo | Descripcion |
|--------|--------|-------------|
| **navegator** | sonnet | Automatizacion web con Go + Chrome DevTools Protocol (chromedp). Perfiles de navegacion |
---
## Parallel Executor
Binario Go en `utils/parallel-executor/` que orquesta la ejecucion paralela de issues usando git worktrees.
### Arquitectura
Sigue el patron **pure core / impure shell** de DevFactory:
```
utils/parallel-executor/
├── core/ # Funciones puras, 0 side effects
│ ├── parser.go # Parsea PARALLEL_EXECUTION_ORDER.md
│ ├── planner.go # Topological sort (Kahn), deteccion de ciclos
│ ├── parser_test.go
│ └── planner_test.go
├── shell/ # Operaciones I/O con Result[T]
│ ├── worktree.go # CRUD de git worktrees
│ ├── executor.go # Invoca `claude -p` en cada worktree
│ └── logger.go # Logs por sesion e issue
├── main.go # CLI
├── go.mod + go.work # DevFactory como dependencia
└── Makefile
```
### Uso
```bash
# Compilar
cd utils/parallel-executor && make build
# Analizar issues y generar plan
./bin/parallel-executor --sort
# Ver que haria sin ejecutar
./bin/parallel-executor --dry-run
# Ejecutar todo
./bin/parallel-executor
# Solo un grupo
./bin/parallel-executor --group 1
# Sin paralelismo
./bin/parallel-executor --sequential
# Solo limpiar worktrees
./bin/parallel-executor --cleanup
```
### Flujo de ejecucion
1. Lee (o genera) `PARALLEL_EXECUTION_ORDER.md`
2. Agrupa issues por dependencias (topological sort)
3. Por cada grupo, crea worktrees en `worktrees/issue-NNNN/`
4. Ejecuta `claude -p` en paralelo dentro de cada worktree
5. Mergea branches exitosas a master (`--no-ff`)
6. Limpia worktrees automaticamente
7. Escribe logs en `logs/`
---
## Stack tecnologico
### Backend (Go)
- **Go 1.22+** con generics
- **DevFactory** (`~/.local_agentes/backend`): Result[T], Option[T], MapSlice, FilterSlice, Reduce, Pipe, Curry
- Patron **pure core / impure shell**: funciones puras en `core/`, I/O wrapeado en Result[T] en `shell/`
- **DuckDB** para analytics, **SQLite** para metadata
- **Bubble Tea** para TUIs
### Frontend (React)
- **React 19** + TypeScript strict + **Vite** + **Tailwind CSS 4** (OKLCH)
- **Frontend_Library** (`~/.local_agentes/frontend`): shadcn/ui, Phosphor Icons, 50+ componentes
- **pnpm** exclusivamente, link via `pnpm add @anthropic/frontend-lib@link:...`
- Vite dedupe obligatorio para `react`, `react-dom`
- **Storybook 10** para documentacion de componentes
### Desktop
- **Wails v2** (Go backend + React frontend)
- Cross-compile: Linux AMD64/ARM64, Windows AMD64
- DevFactory via `go.work`, Frontend_Library via `pnpm link`
### Infraestructura
- **Gitea** self-hosted para repositorios
- **Docker** multi-stage builds
- **Trunk-based development**: ramas `issue/*` y `quick/*`, merge rapido a master
---
## Convenciones
- **Inmutabilidad**: no mutar datos, crear copias nuevas
- **Result[T]** para errores, no `(T, error)` — permite encadenamiento monadico
- **Commits atomicos**: agrupados por tipo (feat, fix, refactor, docs, chore, test)
- **Issues**: formato 3-4 digitos, estado en markdown, sub-issues con sufijo letra
- **Skills con bash scripts**: idempotentes, detectan estado existente, colores en output
BIN
View File
Binary file not shown.
+116
View File
@@ -0,0 +1,116 @@
# 0010 — Consolidar Skills de Issues en una Skill Unificada
## Metadata
| Campo | Valor |
|-------|-------|
| **ID** | 0010 |
| **Estado** | 🟡 pendiente |
| **Prioridad** | media |
| **Tipo** | refactor |
## Dependencias
Ninguna.
**Desbloquea:** mejor mantenibilidad del sistema de skills
---
## Objetivo
Consolidar las 6 skills de gestión de issues (`create-issue`, `auto-create`, `quick-issue`, `fix-issue`, `auto-fix`, `issues-status`) en 2 skills unificadas con flags, reduciendo duplicación y simplificando el mantenimiento.
## Contexto
- Actualmente hay 3 variantes de creación (`create-issue`, `auto-create`, `quick-issue`) que difieren solo en el nivel de confirmación
- Hay 2 variantes de implementación (`fix-issue`, `auto-fix`) que difieren solo en confirmación
- Las 3 skills de análisis/ejecución paralela (`sort-issues`, `parallel-issues`, `execute-parallel`) son un pipeline secuencial que podría ser un solo flujo
- Mantener 6+ skills con lógica casi idéntica genera drift y bugs silenciosos
## Arquitectura
```
skills/
├── issue/SKILL.md — NUEVA: unifica create + auto-create + quick-issue
├── fix/SKILL.md — NUEVA: unifica fix-issue + auto-fix
├── issues-status/SKILL.md — SE MANTIENE (es diferente)
├── parallel-pipeline/SKILL.md — NUEVA: unifica sort + parallel + execute
├── create-issue/ — DEPRECAR
├── auto-create/ — DEPRECAR
├── quick-issue/ — DEPRECAR
├── fix-issue/ — DEPRECAR
├── auto-fix/ — DEPRECAR
├── sort-issues/ — DEPRECAR
├── parallel-issues/ — DEPRECAR
└── execute-parallel/ — DEPRECAR
```
## Tareas
### Fase 1: Skill `/issue` unificada
- [ ] **1.1** Crear `skills/issue/SKILL.md` que acepte flags:
- `/issue` → modo interactivo con confirmación (actual `create-issue`)
- `/issue --auto` → sin confirmación (actual `auto-create`)
- `/issue --quick --text "descripción"` → minimal desde TUI (actual `quick-issue`)
- [ ] **1.2** Extraer lógica común: detección de número, slug, template, git integration
- [ ] **1.3** Tests: verificar los 3 modos producen el mismo resultado
### Fase 2: Skill `/fix` unificada
- [ ] **2.1** Crear `skills/fix/SKILL.md` que acepte flags:
- `/fix <NNNN>` → modo interactivo con confirmación (actual `fix-issue`)
- `/fix <NNNN> --auto` → sin confirmación (actual `auto-fix`)
- [ ] **2.2** Extraer lógica común: lectura de issue, branch, implementación, tests, cierre
### Fase 3: Skill `/parallel` unificada
- [ ] **3.1** Crear `skills/parallel-pipeline/SKILL.md` que combine:
- Análisis de dependencias (actual `sort-issues`)
- Generación de plan paralelo (actual `parallel-issues`)
- Ejecución con worktrees (actual `execute-parallel`)
- [ ] **3.2** Flags: `--dry-run`, `--group N`, `--sequential`, `--plan-only`
### Fase 4: Migración y limpieza
- [ ] **4.1** Mantener skills antiguas como aliases temporales (1 semana)
- [ ] **4.2** Eliminar skills deprecadas
- [ ] **4.3** Actualizar `skills/README.md`
- [ ] **4.4** Actualizar documentación en issues que referencien las skills antiguas
## Ejemplo de uso
```bash
# Antes (3 skills distintas):
/create-issue
/auto-create
/quick-issue --text "bug en parser"
# Después (1 skill con flags):
/issue
/issue --auto
/issue --quick --text "bug en parser"
# Antes (2 skills distintas):
/fix-issue 0010
/auto-fix 0010
# Después:
/fix 0010
/fix 0010 --auto
```
## Decisiones de diseño
- **Flags sobre skills separadas**: Reduce de 8 skills a 3, misma funcionalidad. El flag `--auto` es más intuitivo que recordar `auto-fix` vs `fix-issue`
- **`issues-status` se mantiene separada**: Es un dashboard de consulta, no comparte lógica con creación/implementación
- **Pipeline paralelo como skill única**: sort → plan → execute es siempre secuencial, no tiene sentido invocarlos por separado
## Criterios de aceptación
- [ ] Las 3 nuevas skills cubren 100% de la funcionalidad de las 8 antiguas
- [ ] Los flags son consistentes entre skills (`--auto`, `--dry-run`)
- [ ] Skills antiguas eliminadas sin romper nada
- [ ] README.md de skills actualizado
+6
View File
@@ -21,6 +21,12 @@ Sistema local de issues para trackear mejoras y nuevos agentes.
| 008 | [frontend-lib](008-improve-frontend-lib.md) - Versionado, testing | Media | Pendiente |
| 009 | [gitea](009-improve-gitea.md) - Actions, templates | Media | Pendiente |
## Mejoras a Skills
| # | Issue | Prioridad | Estado |
|---|-------|-----------|--------|
| 010 | [consolidate-issue-skills](010-consolidate-issue-skills.md) - Unificar skills de issues con flags | Media | Pendiente |
## Agentes Completados
| Agente | Descripción | Fecha |
+31
View File
@@ -0,0 +1,31 @@
.PHONY: build test clean install
BIN := parallel-executor
BUILD_DIR := ../../bin
## build: Compila el binario
build:
@mkdir -p $(BUILD_DIR)
go build -trimpath -ldflags="-s -w" -o $(BUILD_DIR)/$(BIN) .
@echo "✓ Built $(BUILD_DIR)/$(BIN)"
## test: Ejecuta tests
test:
go test ./core/... -v
## clean: Elimina artefactos
clean:
rm -f $(BUILD_DIR)/$(BIN)
## install: Copia binario a la skill
install: build
@echo "✓ Binary at $(BUILD_DIR)/$(BIN)"
@echo " Use: $(BUILD_DIR)/$(BIN) --help"
## tidy: go mod tidy
tidy:
go mod tidy
## help: Muestra ayuda
help:
@grep -E '^## ' Makefile | sed 's/## //' | column -t -s ':'
+251
View File
@@ -0,0 +1,251 @@
// Package core contiene funciones puras para el parallel executor.
// Sin side effects: parseo de planes, análisis de dependencias, agrupamiento.
package core
import (
"errors"
"fmt"
"regexp"
"strconv"
"strings"
df "github.com/lucasdataproyects/devfactory/core"
)
// Issue representa una issue parseada del plan de ejecución.
type Issue struct {
Number int
Slug string
Title string
Group int
Deps []int
}
// ExecutionPlan es el plan completo parseado del markdown.
type ExecutionPlan struct {
Groups []IssueGroup
Total int
}
// IssueGroup es un conjunto de issues ejecutables en paralelo.
type IssueGroup struct {
Number int
Name string
Issues []Issue
}
// WorktreeSpec define la especificación para crear un worktree.
type WorktreeSpec struct {
Issue Issue
BranchName string
WorkDir string
}
// ExecutionResult es el resultado de ejecutar una issue.
type ExecutionResult struct {
Issue Issue
Success bool
Duration string
Error string
LogFile string
}
var (
groupHeaderRe = regexp.MustCompile(`^##\s+Grupo\s+(\d+)`)
issueLineRe = regexp.MustCompile(`-\s+Issue\s+#(\d{4})\s*-?\s*(.*)`)
depRe = regexp.MustCompile(`#(\d{4})`)
)
// ParsePlan parsea el contenido markdown de PARALLEL_EXECUTION_ORDER.md.
// Función pura: recibe string, retorna Result[ExecutionPlan].
func ParsePlan(content string) df.Result[ExecutionPlan] {
lines := strings.Split(content, "\n")
if len(lines) == 0 {
return df.Err[ExecutionPlan](errors.New("empty plan"))
}
groups := parsePlanGroups(lines)
if len(groups) == 0 {
return df.Err[ExecutionPlan](errors.New("no groups found in plan"))
}
total := df.Reduce(groups, 0, func(acc int, g IssueGroup) int {
return acc + len(g.Issues)
})
return df.Ok(ExecutionPlan{
Groups: groups,
Total: total,
})
}
// parsePlanGroups extrae los grupos del markdown.
func parsePlanGroups(lines []string) []IssueGroup {
var groups []IssueGroup
var currentGroup *IssueGroup
for _, line := range lines {
trimmed := strings.TrimSpace(line)
if matches := groupHeaderRe.FindStringSubmatch(trimmed); len(matches) > 1 {
if currentGroup != nil {
groups = append(groups, *currentGroup)
}
num, _ := strconv.Atoi(matches[1])
currentGroup = &IssueGroup{
Number: num,
Name: trimmed,
}
continue
}
if currentGroup != nil {
if matches := issueLineRe.FindStringSubmatch(trimmed); len(matches) > 1 {
num, _ := strconv.Atoi(matches[1])
title := strings.TrimSpace(matches[2])
deps := extractDeps(title, num)
issue := Issue{
Number: num,
Slug: issueSlug(num, title),
Title: title,
Group: currentGroup.Number,
Deps: deps,
}
currentGroup.Issues = append(currentGroup.Issues, issue)
}
}
}
if currentGroup != nil && len(currentGroup.Issues) > 0 {
groups = append(groups, *currentGroup)
}
return groups
}
// extractDeps extrae números de issues referenciadas como dependencias.
func extractDeps(text string, selfNum int) []int {
matches := depRe.FindAllStringSubmatch(text, -1)
return df.FilterSlice(
df.MapSlice(matches, func(m []string) int {
n, _ := strconv.Atoi(m[1])
return n
}),
func(n int) bool { return n != selfNum },
)
}
// issueSlug genera el slug de branch para una issue.
func issueSlug(num int, title string) string {
slug := strings.ToLower(title)
slug = regexp.MustCompile(`[^a-z0-9\s-]`).ReplaceAllString(slug, "")
slug = regexp.MustCompile(`\s+`).ReplaceAllString(slug, "-")
slug = regexp.MustCompile(`-{2,}`).ReplaceAllString(slug, "-")
slug = strings.Trim(slug, "-")
words := strings.SplitN(slug, "-", 5)
if len(words) > 4 {
words = words[:4]
}
slug = strings.Join(words, "-")
if slug == "" {
slug = "issue"
}
return formatIssueNum(num) + "-" + slug
}
func formatIssueNum(n int) string {
return fmt.Sprintf("%04d", n)
}
// FilterGroup filtra el plan para ejecutar solo un grupo específico.
func FilterGroup(plan ExecutionPlan, groupNum int) df.Result[ExecutionPlan] {
filtered := df.FilterSlice(plan.Groups, func(g IssueGroup) bool {
return g.Number == groupNum
})
if len(filtered) == 0 {
return df.Err[ExecutionPlan](fmt.Errorf("group not found: %d", groupNum))
}
total := len(filtered[0].Issues)
return df.Ok(ExecutionPlan{Groups: filtered, Total: total})
}
// BuildWorktreeSpecs genera las specs de worktrees para un grupo de issues.
// Función pura: recibe issues y base path, retorna specs.
func BuildWorktreeSpecs(issues []Issue, basePath string) []WorktreeSpec {
return df.MapSlice(issues, func(issue Issue) WorktreeSpec {
branch := "issue/" + issue.Slug
workDir := basePath + "/worktrees/issue-" + formatIssueNum(issue.Number)
return WorktreeSpec{
Issue: issue,
BranchName: branch,
WorkDir: workDir,
}
})
}
// ParseIssueFiles parsea los archivos de issues del directorio dev/issues/.
// Retorna issues con dependencias extraídas del contenido de cada archivo.
func ParseIssueFiles(files map[string]string) []Issue {
var issues []Issue
numRe := regexp.MustCompile(`^(\d{3,4})[a-z]?-(.+)\.md$`)
for filename, content := range files {
matches := numRe.FindStringSubmatch(filename)
if len(matches) < 3 {
continue
}
num, _ := strconv.Atoi(matches[1])
slug := matches[2]
// Extraer título de la primera línea
title := slug
for _, line := range strings.SplitN(content, "\n", 5) {
if strings.HasPrefix(line, "# ") {
title = strings.TrimPrefix(line, "# ")
title = strings.TrimSpace(title)
break
}
}
// Extraer dependencias de "Bloqueada por" o tabla de dependencias
deps := extractAllDeps(content, num)
// Solo incluir si está pendiente
if isIssuePending(content) {
issues = append(issues, Issue{
Number: num,
Slug: slug,
Title: title,
Deps: deps,
})
}
}
return issues
}
func extractAllDeps(content string, selfNum int) []int {
allMatches := depRe.FindAllStringSubmatch(content, -1)
seen := make(map[int]bool)
var deps []int
for _, m := range allMatches {
n, _ := strconv.Atoi(m[1])
if n != selfNum && !seen[n] {
seen[n] = true
deps = append(deps, n)
}
}
return deps
}
func isIssuePending(content string) bool {
lower := strings.ToLower(content)
return strings.Contains(lower, "pendiente") ||
strings.Contains(lower, "🟡") ||
(!strings.Contains(lower, "completado") && !strings.Contains(lower, "✅"))
}
@@ -0,0 +1,89 @@
package core
import (
"testing"
)
func TestParsePlan(t *testing.T) {
plan := `# Plan de Ejecución Paralela
## Grupo 1
- Issue #0003 - testing agent
- Issue #0006 - improve db-reader
## Grupo 2
- Issue #0004 - api client — depende de #0003
`
result := ParsePlan(plan)
if result.IsErr() {
t.Fatalf("ParsePlan failed: %v", result.Error())
}
ep := result.Unwrap()
if len(ep.Groups) != 2 {
t.Errorf("expected 2 groups, got %d", len(ep.Groups))
}
if ep.Total != 3 {
t.Errorf("expected 3 total issues, got %d", ep.Total)
}
if ep.Groups[0].Issues[0].Number != 3 {
t.Errorf("expected issue #0003, got #%04d", ep.Groups[0].Issues[0].Number)
}
if len(ep.Groups[1].Issues[0].Deps) != 1 || ep.Groups[1].Issues[0].Deps[0] != 3 {
t.Errorf("expected issue #0004 to depend on #0003")
}
}
func TestParsePlanEmpty(t *testing.T) {
result := ParsePlan("")
if !result.IsErr() {
t.Error("expected error for empty plan")
}
}
func TestFilterGroup(t *testing.T) {
plan := ExecutionPlan{
Groups: []IssueGroup{
{Number: 1, Issues: []Issue{{Number: 1}}},
{Number: 2, Issues: []Issue{{Number: 2}, {Number: 3}}},
},
Total: 3,
}
result := FilterGroup(plan, 2)
if result.IsErr() {
t.Fatalf("FilterGroup failed: %v", result.Error())
}
filtered := result.Unwrap()
if filtered.Total != 2 {
t.Errorf("expected 2 issues, got %d", filtered.Total)
}
}
func TestFilterGroupNotFound(t *testing.T) {
plan := ExecutionPlan{Groups: []IssueGroup{{Number: 1}}}
result := FilterGroup(plan, 99)
if !result.IsErr() {
t.Error("expected error for non-existent group")
}
}
func TestBuildWorktreeSpecs(t *testing.T) {
issues := []Issue{
{Number: 3, Slug: "0003-testing"},
{Number: 6, Slug: "0006-db-reader"},
}
specs := BuildWorktreeSpecs(issues, "/tmp/project")
if len(specs) != 2 {
t.Fatalf("expected 2 specs, got %d", len(specs))
}
if specs[0].BranchName != "issue/0003-testing" {
t.Errorf("expected branch issue/0003-testing, got %s", specs[0].BranchName)
}
if specs[1].WorkDir != "/tmp/project/worktrees/issue-0006" {
t.Errorf("unexpected workdir: %s", specs[1].WorkDir)
}
}
+217
View File
@@ -0,0 +1,217 @@
package core
import (
"errors"
"fmt"
"sort"
df "github.com/lucasdataproyects/devfactory/core"
)
// TopologicalSort ordena issues por dependencias usando Kahn's algorithm.
// Función pura: retorna Result con los grupos ordenados.
func TopologicalSort(issues []Issue) df.Result[[]IssueGroup] {
if len(issues) == 0 {
return df.Err[[]IssueGroup](errors.New("no issues to sort"))
}
// Construir mapa de issues por número
issueMap := make(map[int]Issue)
for _, issue := range issues {
issueMap[issue.Number] = issue
}
// Calcular in-degree (solo deps que existen en el set)
inDegree := make(map[int]int)
for _, issue := range issues {
if _, exists := inDegree[issue.Number]; !exists {
inDegree[issue.Number] = 0
}
for _, dep := range issue.Deps {
if _, exists := issueMap[dep]; exists {
inDegree[issue.Number]++
}
}
}
// Kahn's algorithm por niveles (cada nivel = un grupo paralelo)
var groups []IssueGroup
resolved := make(map[int]bool)
remaining := len(issues)
groupNum := 1
for remaining > 0 {
// Encontrar issues con in-degree 0
var ready []Issue
for _, issue := range issues {
if !resolved[issue.Number] && inDegree[issue.Number] == 0 {
issue.Group = groupNum
ready = append(ready, issue)
}
}
if len(ready) == 0 {
// Ciclo detectado
var cycleNums []int
for _, issue := range issues {
if !resolved[issue.Number] {
cycleNums = append(cycleNums, issue.Number)
}
}
return df.Err[[]IssueGroup](
fmt.Errorf("circular dependency detected among issues: %v", cycleNums),
)
}
// Ordenar por número dentro del grupo (determinista)
sort.Slice(ready, func(i, j int) bool {
return ready[i].Number < ready[j].Number
})
groups = append(groups, IssueGroup{
Number: groupNum,
Name: fmt.Sprintf("Grupo %d", groupNum),
Issues: ready,
})
// Marcar como resueltas y decrementar in-degree
for _, issue := range ready {
resolved[issue.Number] = true
remaining--
// Decrementar in-degree de dependientes
for _, other := range issues {
if !resolved[other.Number] {
for _, dep := range other.Deps {
if dep == issue.Number {
inDegree[other.Number]--
}
}
}
}
}
groupNum++
}
return df.Ok(groups)
}
// AnalyzeFileConflicts detecta issues que tocan los mismos archivos.
// Recibe un mapa issue_number -> lista de archivos mencionados.
// Retorna pares de issues en conflicto.
func AnalyzeFileConflicts(filesByIssue map[int][]string) []ConflictPair {
var conflicts []ConflictPair
nums := sortedKeys(filesByIssue)
for i := 0; i < len(nums); i++ {
for j := i + 1; j < len(nums); j++ {
shared := intersect(filesByIssue[nums[i]], filesByIssue[nums[j]])
if len(shared) > 0 {
conflicts = append(conflicts, ConflictPair{
IssueA: nums[i],
IssueB: nums[j],
SharedFiles: shared,
})
}
}
}
return conflicts
}
// ConflictPair representa dos issues que comparten archivos.
type ConflictPair struct {
IssueA int
IssueB int
SharedFiles []string
}
// GeneratePlanMarkdown genera el markdown del plan de ejecución.
// Función pura: recibe grupos, retorna string.
func GeneratePlanMarkdown(groups []IssueGroup, conflicts []ConflictPair) string {
var b []byte
b = append(b, "# Plan de Ejecución Paralela\n\n"...)
b = append(b, fmt.Sprintf("Generado automáticamente. Total: %d issues en %d grupos.\n\n",
countIssues(groups), len(groups))...)
if len(conflicts) > 0 {
b = append(b, "## Conflictos Detectados\n\n"...)
for _, c := range conflicts {
b = append(b, fmt.Sprintf("- Issue #%04d y #%04d comparten: %v\n",
c.IssueA, c.IssueB, c.SharedFiles)...)
}
b = append(b, "\n"...)
}
for _, group := range groups {
b = append(b, fmt.Sprintf("## Grupo %d\n\n", group.Number)...)
for _, issue := range group.Issues {
depStr := ""
if len(issue.Deps) > 0 {
depStr = fmt.Sprintf(" — depende de %s", formatDeps(issue.Deps))
}
b = append(b, fmt.Sprintf("- Issue #%04d - %s%s\n", issue.Number, issue.Title, depStr)...)
}
b = append(b, "\n"...)
}
b = append(b, "## Resumen\n\n"...)
b = append(b, "| Métrica | Valor |\n"...)
b = append(b, "|---------|-------|\n"...)
b = append(b, fmt.Sprintf("| Issues totales | %d |\n", countIssues(groups))...)
b = append(b, fmt.Sprintf("| Grupos paralelos | %d |\n", len(groups))...)
if len(groups) > 0 {
sequential := countIssues(groups)
parallel := len(groups)
if sequential > 0 {
saving := ((sequential - parallel) * 100) / sequential
b = append(b, fmt.Sprintf("| Ahorro estimado | %d%% |\n", saving)...)
}
}
return string(b)
}
func countIssues(groups []IssueGroup) int {
return df.Reduce(groups, 0, func(acc int, g IssueGroup) int {
return acc + len(g.Issues)
})
}
func formatDeps(deps []int) string {
strs := df.MapSlice(deps, func(n int) string {
return fmt.Sprintf("#%04d", n)
})
s := ""
for i, str := range strs {
if i > 0 {
s += ", "
}
s += str
}
return s
}
func sortedKeys(m map[int][]string) []int {
keys := make([]int, 0, len(m))
for k := range m {
keys = append(keys, k)
}
sort.Ints(keys)
return keys
}
func intersect(a, b []string) []string {
set := make(map[string]bool)
for _, s := range a {
set[s] = true
}
var result []string
for _, s := range b {
if set[s] {
result = append(result, s)
}
}
return result
}
@@ -0,0 +1,102 @@
package core
import (
"strings"
"testing"
)
func TestTopologicalSort(t *testing.T) {
issues := []Issue{
{Number: 1, Title: "base", Deps: nil},
{Number: 2, Title: "depends on 1", Deps: []int{1}},
{Number: 3, Title: "independent", Deps: nil},
{Number: 4, Title: "depends on 1 and 3", Deps: []int{1, 3}},
}
result := TopologicalSort(issues)
if result.IsErr() {
t.Fatalf("TopologicalSort failed: %v", result.Error())
}
groups := result.Unwrap()
if len(groups) < 2 {
t.Fatalf("expected at least 2 groups, got %d", len(groups))
}
// Group 1 should have issues 1 and 3 (no deps)
g1Nums := make(map[int]bool)
for _, issue := range groups[0].Issues {
g1Nums[issue.Number] = true
}
if !g1Nums[1] || !g1Nums[3] {
t.Errorf("group 1 should contain issues 1 and 3, got %v", groups[0].Issues)
}
}
func TestTopologicalSortCycle(t *testing.T) {
issues := []Issue{
{Number: 1, Title: "a", Deps: []int{2}},
{Number: 2, Title: "b", Deps: []int{1}},
}
result := TopologicalSort(issues)
if !result.IsErr() {
t.Error("expected cycle detection error")
}
if !strings.Contains(result.Error().Error(), "circular") {
t.Errorf("expected circular dependency error, got: %v", result.Error())
}
}
func TestTopologicalSortEmpty(t *testing.T) {
result := TopologicalSort(nil)
if !result.IsErr() {
t.Error("expected error for empty issues")
}
}
func TestAnalyzeFileConflicts(t *testing.T) {
files := map[int][]string{
1: {"core/types.go", "shell/http.go"},
2: {"core/types.go", "app/main.go"},
3: {"app/other.go"},
}
conflicts := AnalyzeFileConflicts(files)
if len(conflicts) != 1 {
t.Fatalf("expected 1 conflict, got %d", len(conflicts))
}
if conflicts[0].IssueA != 1 || conflicts[0].IssueB != 2 {
t.Errorf("expected conflict between 1 and 2, got %d and %d",
conflicts[0].IssueA, conflicts[0].IssueB)
}
if conflicts[0].SharedFiles[0] != "core/types.go" {
t.Errorf("expected shared file core/types.go, got %s", conflicts[0].SharedFiles[0])
}
}
func TestGeneratePlanMarkdown(t *testing.T) {
groups := []IssueGroup{
{Number: 1, Issues: []Issue{
{Number: 1, Title: "base"},
{Number: 3, Title: "other"},
}},
{Number: 2, Issues: []Issue{
{Number: 2, Title: "depends", Deps: []int{1}},
}},
}
md := GeneratePlanMarkdown(groups, nil)
if !strings.Contains(md, "## Grupo 1") {
t.Error("markdown should contain group headers")
}
if !strings.Contains(md, "Issue #0001") {
t.Error("markdown should contain issue references")
}
if !strings.Contains(md, "3 issues") || !strings.Contains(md, "2 grupos") {
// Check summary table
if !strings.Contains(md, "| 3 |") {
t.Error("markdown should contain correct totals")
}
}
}
+23
View File
@@ -0,0 +1,23 @@
module github.com/lucasdataproyects/parallel-executor
go 1.22.2
require github.com/lucasdataproyects/devfactory v0.0.0
require (
github.com/apache/arrow/go/v14 v14.0.2 // indirect
github.com/goccy/go-json v0.10.2 // indirect
github.com/google/flatbuffers v23.5.26+incompatible // indirect
github.com/klauspost/compress v1.16.7 // indirect
github.com/klauspost/cpuid/v2 v2.2.5 // indirect
github.com/marcboeker/go-duckdb v1.6.5 // indirect
github.com/mitchellh/mapstructure v1.5.0 // indirect
github.com/pierrec/lz4/v4 v4.1.18 // indirect
github.com/zeebo/xxh3 v1.0.2 // indirect
golang.org/x/mod v0.13.0 // indirect
golang.org/x/sys v0.13.0 // indirect
golang.org/x/tools v0.14.0 // indirect
golang.org/x/xerrors v0.0.0-20220907171357-04be3eba64a2 // indirect
)
replace github.com/lucasdataproyects/devfactory => /home/lucas/.local_agentes/backend
+45
View File
@@ -0,0 +1,45 @@
github.com/apache/arrow/go/v14 v14.0.2 h1:N8OkaJEOfI3mEZt07BIkvo4sC6XDbL+48MBPWO5IONw=
github.com/apache/arrow/go/v14 v14.0.2/go.mod h1:u3fgh3EdgN/YQ8cVQRguVW3R+seMybFg8QBQ5LU+eBY=
github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
github.com/goccy/go-json v0.10.2 h1:CrxCmQqYDkv1z7lO7Wbh2HN93uovUHgrECaO5ZrCXAU=
github.com/goccy/go-json v0.10.2/go.mod h1:6MelG93GURQebXPDq3khkgXZkazVtN9CRI+MGFi0w8I=
github.com/google/flatbuffers v23.5.26+incompatible h1:M9dgRyhJemaM4Sw8+66GHBu8ioaQmyPLg1b8VwK5WJg=
github.com/google/flatbuffers v23.5.26+incompatible/go.mod h1:1AeVuKshWv4vARoZatz6mlQ0JxURH0Kv5+zNeJKJCa8=
github.com/google/uuid v1.3.1 h1:KjJaJ9iWZ3jOFZIf1Lqf4laDRCasjl0BCmnEGxkdLb4=
github.com/google/uuid v1.3.1/go.mod h1:TIyPZe4MgqvfeYDBFedMoGGpEw/LqOeaOT+nhxU+yHo=
github.com/klauspost/compress v1.16.7 h1:2mk3MPGNzKyxErAw8YaohYh69+pa4sIQSC0fPGCFR9I=
github.com/klauspost/compress v1.16.7/go.mod h1:ntbaceVETuRiXiv4DpjP66DpAtAGkEQskQzEyD//IeE=
github.com/klauspost/cpuid/v2 v2.2.5 h1:0E5MSMDEoAulmXNFquVs//DdoomxaoTY1kUhbc/qbZg=
github.com/klauspost/cpuid/v2 v2.2.5/go.mod h1:Lcz8mBdAVJIBVzewtcLocK12l3Y+JytZYpaMropDUws=
github.com/marcboeker/go-duckdb v1.6.5 h1:XCfR1JVZxsemcSPxRQKK0R0ESfgRMHTEqh3Y+dv40SI=
github.com/marcboeker/go-duckdb v1.6.5/go.mod h1:WtWeqqhZoTke/Nbd7V9lnBx7I2/A/q0SAq/urGzPCMs=
github.com/mitchellh/mapstructure v1.5.0 h1:jeMsZIYE/09sWLaz43PL7Gy6RuMjD2eJVyuac5Z2hdY=
github.com/mitchellh/mapstructure v1.5.0/go.mod h1:bFUtVrKA4DC2yAKiSyO/QUcy7e+RRV2QTWOzhPopBRo=
github.com/pierrec/lz4/v4 v4.1.18 h1:xaKrnTkyoqfh1YItXl56+6KJNVYWlEEPuAQW9xsplYQ=
github.com/pierrec/lz4/v4 v4.1.18/go.mod h1:gZWDp/Ze/IJXGXf23ltt2EXimqmTUXEy0GFuRQyBid4=
github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
github.com/stretchr/testify v1.8.4 h1:CcVxjf3Q8PM0mHUKJCdn+eZZtm5yQwehR5yeSVQQcUk=
github.com/stretchr/testify v1.8.4/go.mod h1:sz/lmYIOXD/1dqDmKjjqLyZ2RngseejIcXlSw2iwfAo=
github.com/zeebo/assert v1.3.0 h1:g7C04CbJuIDKNPFHmsk4hwZDO5O+kntRxzaUoNXj+IQ=
github.com/zeebo/assert v1.3.0/go.mod h1:Pq9JiuJQpG8JLJdtkwrJESF0Foym2/D9XMU5ciN/wJ0=
github.com/zeebo/xxh3 v1.0.2 h1:xZmwmqxHZA8AI603jOQ0tMqmBr9lPeFwGg6d+xy9DC0=
github.com/zeebo/xxh3 v1.0.2/go.mod h1:5NWz9Sef7zIDm2JHfFlcQvNekmcEl9ekUZQQKCYaDcA=
golang.org/x/exp v0.0.0-20231006140011-7918f672742d h1:jtJma62tbqLibJ5sFQz8bKtEM8rJBtfilJ2qTU199MI=
golang.org/x/exp v0.0.0-20231006140011-7918f672742d/go.mod h1:ldy0pHrwJyGW56pPQzzkH36rKxoZW1tw7ZJpeKx+hdo=
golang.org/x/mod v0.13.0 h1:I/DsJXRlw/8l/0c24sM9yb0T4z9liZTduXvdAWYiysY=
golang.org/x/mod v0.13.0/go.mod h1:hTbmBsO62+eylJbnUtE2MGJUyE7QWk4xUqPFrRgJ+7c=
golang.org/x/sync v0.4.0 h1:zxkM55ReGkDlKSM+Fu41A+zmbZuaPVbGMzvvdUPznYQ=
golang.org/x/sync v0.4.0/go.mod h1:FU7BRWz2tNW+3quACPkgCx/L+uEAv1htQ0V83Z9Rj+Y=
golang.org/x/sys v0.5.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
golang.org/x/sys v0.13.0 h1:Af8nKPmuFypiUBjVoU9V20FiaFXOcuZI21p0ycVYYGE=
golang.org/x/sys v0.13.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
golang.org/x/tools v0.14.0 h1:jvNa2pY0M4r62jkRQ6RwEZZyPcymeL9XZMLBbV7U2nc=
golang.org/x/tools v0.14.0/go.mod h1:uYBEerGOWcJyEORxN+Ek8+TT266gXkNlHdJBwexUsBg=
golang.org/x/xerrors v0.0.0-20220907171357-04be3eba64a2 h1:H2TDz8ibqkAF6YGhCdN3jS9O0/s90v0rJh3X/OLHEUk=
golang.org/x/xerrors v0.0.0-20220907171357-04be3eba64a2/go.mod h1:K8+ghG5WaK9qNqU5K3HdILfMLy1f3aNYFI/wnl100a8=
gonum.org/v1/gonum v0.12.0 h1:xKuo6hzt+gMav00meVPUlXwSdoEJP46BR+wdxQEFK2o=
gonum.org/v1/gonum v0.12.0/go.mod h1:73TDxJfAAHeA8Mk9mf8NlIppyhQNo5GLTcYeqgo2lvY=
gopkg.in/yaml.v3 v3.0.1 h1:fxVm/GzAzEWqLHuvctI91KS9hhNmmWOoWu0XTYJS7CA=
gopkg.in/yaml.v3 v3.0.1/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
+6
View File
@@ -0,0 +1,6 @@
go 1.22.2
use (
.
/home/lucas/.local_agentes/backend
)
+352
View File
@@ -0,0 +1,352 @@
// parallel-executor — Orquestador de ejecución paralela de issues.
//
// Crea git worktrees, ejecuta claude en cada uno, y mergea los resultados.
// Usa DevFactory para Result[T], MapSlice, y operaciones I/O.
//
// Uso:
//
// parallel-executor [flags]
// --plan <file> Plan markdown (default: PARALLEL_EXECUTION_ORDER.md)
// --group <N> Ejecutar solo grupo N
// --sequential Sin paralelismo
// --dry-run Solo mostrar plan sin ejecutar
// --timeout <min> Timeout por issue en minutos (default: 30)
// --cleanup Solo limpiar worktrees existentes
// --sort Analizar issues y generar plan (no ejecutar)
package main
import (
"flag"
"fmt"
"os"
"sync"
"time"
dfshell "github.com/lucasdataproyects/devfactory/shell"
pcore "github.com/lucasdataproyects/parallel-executor/core"
"github.com/lucasdataproyects/parallel-executor/shell"
)
// CLI colors
const (
red = "\033[0;31m"
green = "\033[0;32m"
yellow = "\033[1;33m"
blue = "\033[0;34m"
cyan = "\033[0;36m"
nc = "\033[0m"
)
func main() {
planFile := flag.String("plan", "PARALLEL_EXECUTION_ORDER.md", "plan markdown file")
groupNum := flag.Int("group", 0, "execute only this group number")
sequential := flag.Bool("sequential", false, "disable parallel execution")
dryRun := flag.Bool("dry-run", false, "show plan without executing")
timeoutMin := flag.Int("timeout", 30, "timeout per issue in minutes")
cleanup := flag.Bool("cleanup", false, "only cleanup existing worktrees")
sortMode := flag.Bool("sort", false, "analyze issues and generate plan")
flag.Parse()
repoRoot, err := os.Getwd()
if err != nil {
fatal("cannot get working directory: %v", err)
}
// --- Modo cleanup ---
if *cleanup {
info("Cleaning up worktrees...")
result := shell.CleanupAllWorktrees(repoRoot)
if result.IsErr() {
fatal("cleanup failed: %v", result.Error())
}
ok("Cleaned %d worktrees", result.Unwrap())
return
}
// --- Modo sort: analizar issues y generar plan ---
if *sortMode {
generatePlan(repoRoot)
return
}
// --- Ejecutar plan ---
executePlan(repoRoot, *planFile, *groupNum, *sequential, *dryRun, *timeoutMin)
}
func generatePlan(repoRoot string) {
issuesDir := repoRoot + "/dev/issues"
if !dfshell.DirExists(issuesDir) {
fatal("issues directory not found: %s", issuesDir)
}
info("Reading issues from %s...", issuesDir)
filesResult := shell.ReadIssueFiles(issuesDir)
if filesResult.IsErr() {
fatal("cannot read issues: %v", filesResult.Error())
}
issues := pcore.ParseIssueFiles(filesResult.Unwrap())
if len(issues) == 0 {
warn("No pending issues found")
return
}
info("Found %d pending issues", len(issues))
// Topological sort
groupsResult := pcore.TopologicalSort(issues)
if groupsResult.IsErr() {
fatal("dependency analysis failed: %v", groupsResult.Error())
}
groups := groupsResult.Unwrap()
// Generate markdown
markdown := pcore.GeneratePlanMarkdown(groups, nil)
outFile := repoRoot + "/PARALLEL_EXECUTION_ORDER.md"
writeResult := dfshell.WriteString(outFile, markdown)
if writeResult.IsErr() {
fatal("cannot write plan: %v", writeResult.Error())
}
ok("Plan generated: %s", outFile)
fmt.Printf("\n%sIssues:%s %d\n", cyan, nc, len(issues))
fmt.Printf("%sGroups:%s %d\n", cyan, nc, len(groups))
for _, g := range groups {
fmt.Printf(" %sGrupo %d:%s %d issues\n", blue, g.Number, nc, len(g.Issues))
for _, issue := range g.Issues {
fmt.Printf(" - #%04d %s\n", issue.Number, issue.Title)
}
}
}
func executePlan(repoRoot string, planFile string, groupNum int, sequential bool, dryRun bool, timeoutMin int) {
// Leer plan
planContent := dfshell.ReadString(planFile)
if planContent.IsErr() {
// Intentar generar plan automáticamente
warn("Plan not found, generating from issues...")
generatePlan(repoRoot)
planContent = dfshell.ReadString(planFile)
if planContent.IsErr() {
fatal("cannot read plan: %v", planContent.Error())
}
}
planResult := pcore.ParsePlan(planContent.Unwrap())
if planResult.IsErr() {
fatal("cannot parse plan: %v", planResult.Error())
}
plan := planResult.Unwrap()
// Filtrar por grupo si se especificó
if groupNum > 0 {
filtered := pcore.FilterGroup(plan, groupNum)
if filtered.IsErr() {
fatal("group filter: %v", filtered.Error())
}
plan = filtered.Unwrap()
}
info("Plan: %d issues in %d groups", plan.Total, len(plan.Groups))
// --- Dry run: solo mostrar ---
if dryRun {
for _, group := range plan.Groups {
fmt.Printf("\n%s%s%s (%d issues)\n", cyan, group.Name, nc, len(group.Issues))
specs := pcore.BuildWorktreeSpecs(group.Issues, repoRoot)
for _, spec := range specs {
fmt.Printf(" #%04d → %s\n", spec.Issue.Number, spec.BranchName)
fmt.Printf(" %s\n", spec.WorkDir)
}
}
fmt.Printf("\n%sDry run complete. No worktrees created.%s\n", yellow, nc)
return
}
// --- Logger ---
loggerResult := shell.NewLogger(repoRoot)
if loggerResult.IsErr() {
fatal("cannot create logger: %v", loggerResult.Error())
}
logger := loggerResult.Unwrap()
info("Logs: %s", logger.SessionLogFile())
// --- Ejecutar grupos secuencialmente, issues dentro de cada grupo en paralelo ---
timeout := time.Duration(timeoutMin) * time.Minute
var allResults []pcore.ExecutionResult
for _, group := range plan.Groups {
fmt.Printf("\n%s══════════════════════════════════════%s\n", cyan, nc)
fmt.Printf("%s %s — %d issues%s\n", cyan, group.Name, len(group.Issues), nc)
fmt.Printf("%s══════════════════════════════════════%s\n", cyan, nc)
specs := pcore.BuildWorktreeSpecs(group.Issues, repoRoot)
// Crear worktrees
info("Creating %d worktrees...", len(specs))
var validSpecs []pcore.WorktreeSpec
for _, spec := range specs {
result := shell.CreateWorktree(spec, repoRoot)
if result.IsErr() {
warn("Failed to create worktree for #%04d: %v", spec.Issue.Number, result.Error())
allResults = append(allResults, pcore.ExecutionResult{
Issue: spec.Issue,
Success: false,
Error: result.Error().Error(),
})
continue
}
ok("Worktree: %s → %s", spec.BranchName, result.Unwrap())
validSpecs = append(validSpecs, spec)
}
// Ejecutar
var groupResults []pcore.ExecutionResult
if sequential || len(validSpecs) == 1 {
groupResults = executeSequential(validSpecs, timeout, logger)
} else {
groupResults = executeParallel(validSpecs, timeout, logger)
}
allResults = append(allResults, groupResults...)
// Mergear las exitosas
for _, r := range groupResults {
if !r.Success {
continue
}
spec := findSpec(validSpecs, r.Issue.Number)
if spec == nil {
continue
}
info("Merging %s...", spec.BranchName)
mergeResult := shell.MergeBranchToMaster(spec.BranchName, repoRoot)
if mergeResult.IsErr() {
warn("Merge failed for %s: %v", spec.BranchName, mergeResult.Error())
} else {
ok("Merged %s", spec.BranchName)
}
}
// Limpiar worktrees del grupo
for _, spec := range validSpecs {
shell.RemoveWorktree(spec, repoRoot)
}
}
// --- Resumen ---
summaryResult := logger.WriteSummary(allResults)
summaryFile := ""
if summaryResult.IsOk() {
summaryFile = summaryResult.Unwrap()
}
fmt.Printf("\n%s══════════════════════════════════════%s\n", green, nc)
fmt.Printf("%s Execution Complete%s\n", green, nc)
fmt.Printf("%s══════════════════════════════════════%s\n\n", green, nc)
succeeded := 0
failed := 0
for _, r := range allResults {
if r.Success {
succeeded++
fmt.Printf(" %s✓%s #%04d %s (%s)\n", green, nc, r.Issue.Number, r.Issue.Title, r.Duration)
} else {
failed++
fmt.Printf(" %s✗%s #%04d %s — %s\n", red, nc, r.Issue.Number, r.Issue.Title, r.Error)
}
}
fmt.Printf("\n Total: %d | Succeeded: %s%d%s | Failed: %s%d%s\n",
len(allResults), green, succeeded, nc, red, failed, nc)
if summaryFile != "" {
fmt.Printf(" Summary: %s\n", summaryFile)
}
fmt.Printf(" Log: %s\n", logger.SessionLogFile())
// Cleanup final
shell.CleanupAllWorktrees(repoRoot)
if failed > 0 {
os.Exit(1)
}
}
func executeSequential(specs []pcore.WorktreeSpec, timeout time.Duration, logger *shell.Logger) []pcore.ExecutionResult {
results := make([]pcore.ExecutionResult, 0, len(specs))
for _, spec := range specs {
info("Executing #%04d %s...", spec.Issue.Number, spec.Issue.Title)
logger.LogIssueStart(spec.Issue)
result := shell.ExecuteIssue(spec, timeout)
logger.LogIssueResult(result)
if result.Success {
ok("#%04d completed in %s", spec.Issue.Number, result.Duration)
} else {
warn("#%04d failed: %s", spec.Issue.Number, result.Error)
}
results = append(results, result)
}
return results
}
func executeParallel(specs []pcore.WorktreeSpec, timeout time.Duration, logger *shell.Logger) []pcore.ExecutionResult {
results := make([]pcore.ExecutionResult, len(specs))
var wg sync.WaitGroup
for i, spec := range specs {
wg.Add(1)
go func(idx int, s pcore.WorktreeSpec) {
defer wg.Done()
info("[goroutine] Executing #%04d %s...", s.Issue.Number, s.Issue.Title)
logger.LogIssueStart(s.Issue)
result := shell.ExecuteIssue(s, timeout)
logger.LogIssueResult(result)
results[idx] = result
if result.Success {
ok("[goroutine] #%04d completed in %s", s.Issue.Number, result.Duration)
} else {
warn("[goroutine] #%04d failed: %s", s.Issue.Number, result.Error)
}
}(i, spec)
}
wg.Wait()
return results
}
func findSpec(specs []pcore.WorktreeSpec, issueNum int) *pcore.WorktreeSpec {
for _, s := range specs {
if s.Issue.Number == issueNum {
return &s
}
}
return nil
}
func info(format string, args ...any) {
fmt.Printf("%s[INFO]%s %s\n", blue, nc, fmt.Sprintf(format, args...))
}
func ok(format string, args ...any) {
fmt.Printf("%s[OK]%s %s\n", green, nc, fmt.Sprintf(format, args...))
}
func warn(format string, args ...any) {
fmt.Printf("%s[WARN]%s %s\n", yellow, nc, fmt.Sprintf(format, args...))
}
func fatal(format string, args ...any) {
fmt.Fprintf(os.Stderr, "%s[ERROR]%s %s\n", red, nc, fmt.Sprintf(format, args...))
os.Exit(1)
}
+127
View File
@@ -0,0 +1,127 @@
package shell
import (
"fmt"
"strings"
"time"
"github.com/lucasdataproyects/devfactory/core"
dfshell "github.com/lucasdataproyects/devfactory/shell"
pcore "github.com/lucasdataproyects/parallel-executor/core"
)
const defaultTimeout = 30 * time.Minute
// ExecuteIssue ejecuta claude para resolver una issue en un worktree.
// Invoca `claude -p` con el prompt de fix-issue dentro del worktree.
func ExecuteIssue(spec pcore.WorktreeSpec, timeout time.Duration) pcore.ExecutionResult {
start := time.Now()
if timeout == 0 {
timeout = defaultTimeout
}
prompt := fmt.Sprintf(
"Read the issue file dev/issues/%04d-*.md and implement all tasks. "+
"Run tests after each change. Follow pure core / impure shell pattern. "+
"When done, commit all changes with a descriptive message.",
spec.Issue.Number,
)
result := dfshell.RunWithTimeout("claude",
timeout,
"-p", prompt,
"--cwd", spec.WorkDir,
)
duration := time.Since(start).Round(time.Second).String()
if result.IsErr() {
return pcore.ExecutionResult{
Issue: spec.Issue,
Success: false,
Duration: duration,
Error: result.Error().Error(),
}
}
cmdResult := result.Unwrap()
if !cmdResult.Success() {
return pcore.ExecutionResult{
Issue: spec.Issue,
Success: false,
Duration: duration,
Error: cmdResult.Stderr,
}
}
return pcore.ExecutionResult{
Issue: spec.Issue,
Success: true,
Duration: duration,
}
}
// PushWorktreeBranch hace push de la branch del worktree al remote.
func PushWorktreeBranch(spec pcore.WorktreeSpec) core.Result[struct{}] {
result := dfshell.Run("git", "-C", spec.WorkDir,
"push", "-u", "origin", spec.BranchName,
)
if result.IsErr() {
return core.Err[struct{}](fmt.Errorf("push failed for %s: %w",
spec.BranchName, result.Error()))
}
return core.Ok(struct{}{})
}
// MergeBranchToMaster mergea una branch a master con --no-ff.
func MergeBranchToMaster(branchName string, repoRoot string) core.Result[struct{}] {
// Checkout master
result := dfshell.Run("git", "-C", repoRoot, "checkout", "master")
if result.IsErr() {
return core.Err[struct{}](result.Error())
}
// Merge --no-ff
message := fmt.Sprintf("merge: %s — parallel execution", branchName)
result = dfshell.Run("git", "-C", repoRoot,
"merge", "--no-ff", "-m", message, branchName,
)
if result.IsErr() {
return core.Err[struct{}](fmt.Errorf("merge failed for %s: %w",
branchName, result.Error()))
}
return core.Ok(struct{}{})
}
// DeleteBranch elimina una branch local.
func DeleteBranch(branchName string, repoRoot string) core.Result[struct{}] {
result := dfshell.Run("git", "-C", repoRoot, "branch", "-d", branchName)
if result.IsErr() {
return core.Err[struct{}](result.Error())
}
return core.Ok(struct{}{})
}
// ReadIssueFiles lee todos los archivos de issues de un directorio.
func ReadIssueFiles(issuesDir string) core.Result[map[string]string] {
entries := dfshell.ListDir(issuesDir)
if entries.IsErr() {
return core.Err[map[string]string](entries.Error())
}
files := make(map[string]string)
for _, entry := range entries.Unwrap() {
if !strings.HasSuffix(entry, ".md") || entry == "README.md" {
continue
}
content := dfshell.ReadString(issuesDir + "/" + entry)
if content.IsOk() {
files[entry] = content.Unwrap()
}
}
return core.Ok(files)
}
+135
View File
@@ -0,0 +1,135 @@
package shell
import (
"fmt"
"strings"
"time"
"github.com/lucasdataproyects/devfactory/core"
dfshell "github.com/lucasdataproyects/devfactory/shell"
pcore "github.com/lucasdataproyects/parallel-executor/core"
)
// Logger escribe logs de la ejecución paralela a disco.
type Logger struct {
logsDir string
sessionID string
}
// NewLogger crea un logger para la sesión actual.
func NewLogger(repoRoot string) core.Result[*Logger] {
sessionID := time.Now().Format("20060102-150405")
logsDir := repoRoot + "/logs"
result := dfshell.MkdirAll(logsDir)
if result.IsErr() {
return core.Err[*Logger](result.Error())
}
return core.Ok(&Logger{
logsDir: logsDir,
sessionID: sessionID,
})
}
// LogIssueStart registra el inicio de ejecución de una issue.
func (l *Logger) LogIssueStart(issue pcore.Issue) {
msg := fmt.Sprintf("[%s] START Issue #%04d - %s\n",
time.Now().Format("15:04:05"), issue.Number, issue.Title)
l.appendToSession(msg)
}
// LogIssueResult registra el resultado de una issue.
func (l *Logger) LogIssueResult(result pcore.ExecutionResult) {
status := "SUCCESS"
if !result.Success {
status = "FAILED"
}
msg := fmt.Sprintf("[%s] %s Issue #%04d - %s (duration: %s)",
time.Now().Format("15:04:05"), status,
result.Issue.Number, result.Issue.Title, result.Duration)
if result.Error != "" {
msg += "\n Error: " + result.Error
}
msg += "\n"
l.appendToSession(msg)
// Log individual por issue
issueLogFile := fmt.Sprintf("%s/issue-%04d-%s.log",
l.logsDir, result.Issue.Number, l.sessionID)
content := fmt.Sprintf("Issue #%04d - %s\nStatus: %s\nDuration: %s\n",
result.Issue.Number, result.Issue.Title, status, result.Duration)
if result.Error != "" {
content += "Error:\n" + result.Error + "\n"
}
dfshell.WriteString(issueLogFile, content)
}
// WriteSummary escribe el resumen consolidado.
func (l *Logger) WriteSummary(results []pcore.ExecutionResult) core.Result[string] {
summaryFile := fmt.Sprintf("%s/summary-%s.txt", l.logsDir, l.sessionID)
var b strings.Builder
b.WriteString("=" + strings.Repeat("=", 59) + "\n")
b.WriteString(fmt.Sprintf(" Parallel Execution Summary — %s\n", l.sessionID))
b.WriteString("=" + strings.Repeat("=", 59) + "\n\n")
succeeded := 0
failed := 0
for _, r := range results {
if r.Success {
succeeded++
} else {
failed++
}
}
b.WriteString(fmt.Sprintf("Total: %d\n", len(results)))
b.WriteString(fmt.Sprintf("Succeeded: %d\n", succeeded))
b.WriteString(fmt.Sprintf("Failed: %d\n\n", failed))
b.WriteString("Results:\n")
b.WriteString(strings.Repeat("-", 60) + "\n")
for _, r := range results {
status := "✓"
if !r.Success {
status = "✗"
}
b.WriteString(fmt.Sprintf(" %s #%04d %-30s %s\n",
status, r.Issue.Number, r.Issue.Title, r.Duration))
if r.Error != "" {
b.WriteString(fmt.Sprintf(" Error: %s\n", truncate(r.Error, 80)))
}
}
b.WriteString(strings.Repeat("-", 60) + "\n")
writeResult := dfshell.WriteString(summaryFile, b.String())
if writeResult.IsErr() {
return core.Err[string](writeResult.Error())
}
return core.Ok(summaryFile)
}
// SessionLogFile retorna la ruta del log de sesión.
func (l *Logger) SessionLogFile() string {
return fmt.Sprintf("%s/parallel-execution-%s.log", l.logsDir, l.sessionID)
}
func (l *Logger) appendToSession(msg string) {
logFile := l.SessionLogFile()
dfshell.AppendFile(logFile, []byte(msg))
}
func truncate(s string, max int) string {
if len(s) <= max {
return s
}
return s[:max-3] + "..."
}
+134
View File
@@ -0,0 +1,134 @@
// Package shell contiene operaciones I/O del parallel executor.
// Todas las funciones retornan Result[T] y tienen side effects (git, filesystem).
package shell
import (
"context"
"fmt"
"os"
"strings"
"time"
"github.com/lucasdataproyects/devfactory/core"
dfshell "github.com/lucasdataproyects/devfactory/shell"
pcore "github.com/lucasdataproyects/parallel-executor/core"
)
// CreateWorktree crea un git worktree para una issue.
// Crea branch desde master y configura el directorio de trabajo.
func CreateWorktree(spec pcore.WorktreeSpec, repoRoot string) core.Result[string] {
// Verificar que no existe ya
if dfshell.DirExists(spec.WorkDir) {
return core.Ok(spec.WorkDir)
}
// Crear directorio padre si no existe
parentDir := spec.WorkDir[:strings.LastIndex(spec.WorkDir, "/")]
mkResult := dfshell.MkdirAll(parentDir)
if mkResult.IsErr() {
return core.Err[string](mkResult.Error())
}
// Actualizar master primero
ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
defer cancel()
updateResult := dfshell.RunWithContext(ctx, "git", "-C", repoRoot, "fetch", "origin", "master")
if updateResult.IsErr() {
// No fatal — puede no tener remote
}
// Crear worktree con nueva branch desde master
result := dfshell.Run("git", "-C", repoRoot,
"worktree", "add",
"-b", spec.BranchName,
spec.WorkDir,
"master",
)
if result.IsErr() {
// Branch puede existir — intentar sin -b
result = dfshell.Run("git", "-C", repoRoot,
"worktree", "add",
spec.WorkDir,
spec.BranchName,
)
if result.IsErr() {
return core.Err[string](fmt.Errorf("failed to create worktree for issue #%04d: %w",
spec.Issue.Number, result.Error()))
}
}
return core.Ok(spec.WorkDir)
}
// RemoveWorktree elimina un worktree y su branch.
func RemoveWorktree(spec pcore.WorktreeSpec, repoRoot string) core.Result[struct{}] {
if !dfshell.DirExists(spec.WorkDir) {
return core.Ok(struct{}{})
}
// Remover worktree
result := dfshell.Run("git", "-C", repoRoot,
"worktree", "remove", "--force", spec.WorkDir,
)
if result.IsErr() {
// Fallback: eliminar directorio manualmente
os.RemoveAll(spec.WorkDir)
// Prune worktrees huérfanos
dfshell.Run("git", "-C", repoRoot, "worktree", "prune")
}
return core.Ok(struct{}{})
}
// CleanupAllWorktrees limpia todos los worktrees del directorio worktrees/.
func CleanupAllWorktrees(repoRoot string) core.Result[int] {
worktreesDir := repoRoot + "/worktrees"
if !dfshell.DirExists(worktreesDir) {
return core.Ok(0)
}
entries := dfshell.ListDir(worktreesDir)
if entries.IsErr() {
return core.Ok(0)
}
count := 0
for _, entry := range entries.Unwrap() {
fullPath := worktreesDir + "/" + entry
dfshell.Run("git", "-C", repoRoot, "worktree", "remove", "--force", fullPath)
count++
}
// Prune
dfshell.Run("git", "-C", repoRoot, "worktree", "prune")
// Eliminar directorio vacío
os.Remove(worktreesDir)
return core.Ok(count)
}
// ListWorktrees devuelve los worktrees activos.
func ListWorktrees(repoRoot string) core.Result[[]string] {
result := dfshell.Run("git", "-C", repoRoot, "worktree", "list", "--porcelain")
if result.IsErr() {
return core.Err[[]string](result.Error())
}
output := result.Unwrap().Output()
lines := strings.Split(output, "\n")
var paths []string
for _, line := range lines {
if strings.HasPrefix(line, "worktree ") {
path := strings.TrimPrefix(line, "worktree ")
if path != repoRoot {
paths = append(paths, path)
}
}
}
return core.Ok(paths)
}