prjct-cli 0.8.6 → 0.9.1

package/templates/commands/ask.md

@@ -7,380 +7,67 @@ description: 'Conversational intent to action translator - helps users understan
 
 ## Purpose
 
-Translate natural language intent into actionable prjct command flows. Helps users who know WHAT they want but don't know HOW to do it with prjct.
-
-## Usage
-
-```
-/p:ask "<what you want to do>"
-```
+Translate natural language intent into actionable prjct command flows. Helps users who know WHAT they want but don't know HOW.
 
 ## Flow
 
 1. **Understand intent**: Parse user's natural language description
-2. **Analyze context**: Read project state (core/now.md, core/next.md, planning/roadmap.md)
+2. **Analyze context**: Read project state (`core/now.md`, `core/next.md`, `planning/roadmap.md`)
 3. **Recommend flow**: Suggest specific command sequence with explanations
-4. **Offer templates**: If applicable, show examples from similar use cases
-5. **Ask confirmation**: Interactive - don't execute automatically
+4. **Ask confirmation**: Interactive - don't execute automatically
 
 ## Intent Categories
 
-Claude analyzes and maps to one of these patterns:
-
-### 1. Feature Development
-Keywords: "add", "implement", "create", "build", "agregar", "implementar", "crear"
+**1. Feature Development**: "add", "implement", "create", "build"
+→ `/p:feature "{desc}"` → `/p:done` (iterate) → `/p:ship "{name}"`
 
-**Recommended flow:**
-```
-1. /p:feature "{description}"
-   → Value analysis (impact/effort/timing)
-   → Task breakdown
-   → Auto-start first task
-
-2. /p:done (after each task)
-
-3. /p:ship "{feature name}" (when complete)
-```
+**2. Performance/Optimization**: "optimize", "improve", "faster", "memory leak"
+→ `/p:feature "optimize {area}"` → `/p:build 1` → `/p:done` → `/p:ship "optimization"`
 
-### 2. Performance/Optimization
-Keywords: "optimize", "improve", "faster", "performance", "memory leak", "optimizar", "mejorar"
+**3. Bug Fixing**: "bug", "error", "fix", "broken", "not working"
+→ `/p:bug "{desc}"` → `/p:build "{bug task}"` → `/p:done` → `/p:ship "bug fixes"`
 
-**Recommended flow:**
-```
-1. /p:feature "optimize {area}"
-   → Break down into measurable tasks:
-     • Profile and identify bottlenecks
-     • Implement specific optimizations
-     • Measure improvements
-     • Document findings
-
-2. /p:build 1 (start with profiling)
-
-3. /p:done (iterate through tasks)
-
-4. /p:ship "performance optimization"
-```
-
-### 3. Bug Fixing
-Keywords: "bug", "error", "fix", "broken", "not working", "arreglar", "error"
-
-**Recommended flow:**
-```
-1. /p:bug "{description}"
-   → Auto-prioritized based on severity
-   → Added to queue
+**4. Design/Architecture**: "design", "architecture", "structure", "plan"
+→ `/p:design {target} --type {architecture|api|component|database|flow}` → `/p:feature "implement {design}"` → `/p:build 1`
 
-2. /p:build "{bug task}" (if high priority)
-
-3. /p:done (when fixed)
-
-4. /p:ship "bug fixes" (if part of larger batch)
-```
-
-### 4. Design/Architecture
-Keywords: "design", "architecture", "structure", "plan", "diseñar", "arquitectura"
-
-**Recommended flow:**
-```
-1. /p:design {target} --type {architecture|api|component|database|flow}
-   → Create design document
-   → Plan implementation
-
-2. /p:feature "implement {design}"
-   → Convert design to tasks
-
-3. /p:build 1 (start implementation)
-```
-
-### 5. Lost/Confused
-Keywords: "don't know", "help", "stuck", "what should", "no sé", "ayuda"
-
-**Recommended flow:**
-```
-1. /p:suggest
-   → Context-aware recommendations
-   → Shows current state
-   → Suggests next best action
-
-OR
-
-2. /p:help
-   → Interactive guide
-   → Shows available commands
-   → Examples
-```
+**5. Lost/Confused**: "don't know", "help", "stuck", "what should"
+→ `/p:suggest` (context-aware recommendations) OR `/p:help` (interactive guide)
 
 ## Response Format
 
 ```
 ━━━━━━━━━━━━━━━━━━━━━━━━━━━
-🎯 Entiendo: {interpreted_intent}
+🎯 I understand: {interpreted_intent}
 ━━━━━━━━━━━━━━━━━━━━━━━━━━━
 
-📊 TU CONTEXTO:
-  • Estado actual: {current_state}
-  • Tareas activas: {active_tasks}
-  • Último ship: {time_ago}
+📊 YOUR CONTEXT:
+  • Current state: {current_state}
+  • Active tasks: {active_tasks}
+  • Last ship: {time_ago}
 
 ━━━━━━━━━━━━━━━━━━━━━━━━━━━
 
-💡 RECOMENDACIÓN DE FLUJO:
+💡 RECOMMENDED FLOW:
 
 {step_by_step_command_flow}
 
-Cada paso explicado:
+Each step explained:
 → {command_1}: {why_this_command}
 → {command_2}: {what_it_does}
 → {command_3}: {expected_outcome}
 
 ━━━━━━━━━━━━━━━━━━━━━━━━━━━
 
-✨ EJEMPLOS SIMILARES:
-
-"{similar_example_1}"
-"{similar_example_2}"
-
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-¿Quieres que ejecute el paso 1?
-[Sí] [No, solo dame las tareas] [Necesito más ayuda]
-```
-
-## Examples
-
-### Example 1: Performance Optimization
-
-**Input:**
-```
-/p:ask "quiero mejorar el performance y evitar memory leaks"
-```
-
-**Output:**
-```
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-🎯 Entiendo: optimización de performance + solución de memory leaks
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-📊 TU CONTEXTO:
-  • Estado actual: sin tarea activa
-  • Tareas en cola: 3
-  • Último ship: hace 2 días
-
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-💡 RECOMENDACIÓN DE FLUJO:
-
-1. /p:feature "optimizar performance y memory leaks"
-
-   Claude analizará y creará tareas como:
-   • Setup profiler (Chrome DevTools/React DevTools)
-   • Identificar memory leaks con heap snapshots
-   • Optimizar re-renders innecesarios
-   • Implementar memoization (useMemo, useCallback)
-   • Code splitting para reducir bundle size
-   • Medir mejoras y documentar
-
-2. /p:build 1
-   → Empezar con el profiling (necesitas datos primero)
-
-3. /p:done
-   → Después de cada tarea completada
-
-4. /p:ship "performance optimization"
-   → Cuando hayas terminado todas las mejoras
-
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-✨ POR QUÉ ESTE FLUJO:
-
-→ /p:feature: Analiza el valor (impacto alto, esfuerzo medio)
-             Recomienda timing (hacer ahora antes de más features)
-             Breakdown inteligente (primero medir, luego optimizar)
-
-→ /p:build:   Asigna agente apropiado (probablemente FE)
-             Tracking de tiempo
-             Focus en una tarea a la vez
-
-→ /p:done:    Registra progreso
-             Mantiene momentum
-             Mueve a siguiente tarea
-
-→ /p:ship:    Commit + push + celebración
-             Actualiza docs
-             Registra el logro
-
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-¿Empezamos con /p:feature para analizar el valor y crear las tareas?
-```
-
-### Example 2: New Feature
-
-**Input:**
-```
-/p:ask "I want to add user authentication"
+Would you like me to execute step 1?
 ```
 
-**Output:**
-```
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-🎯 I understand: implement user authentication system
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-📊 YOUR CONTEXT:
-  • Current state: working on "navbar redesign"
-  • Queue: 2 tasks
-  • Last ship: 1 day ago
-
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-⚠️  IMPORTANTE: Tienes tarea activa
-
-Opciones:
-
-A) TERMINAR PRIMERO (Recomendado)
-   1. /p:done → Terminar "navbar redesign"
-   2. /p:feature "user authentication"
-   3. /p:build 1 → Empezar auth
-
-B) GUARDAR PARA DESPUÉS
-   1. /p:feature "user authentication"
-      → Se agrega a la cola
-   2. Continúa con navbar
-   3. /p:done cuando termines navbar
-   4. /p:build "authentication" cuando estés listo
-
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-💡 RECOMENDACIÓN:
-
-Termina navbar primero (mantiene focus)
-Luego /p:feature "authentication" hará:
-
-• Análisis de valor (impacto: HIGH)
-• Stack analysis (sugiere: Clerk, Auth0, o custom JWT)
-• Task breakdown:
-  - Setup auth provider
-  - Implement login/signup UI
-  - Protected routes
-  - Session management
-  - Testing
-
-• Auto-start primera tarea
-
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-¿Qué prefieres: A) Terminar navbar primero, o B) Agregar a cola?
-```
-
-### Example 3: Lost/Confused
-
-**Input:**
-```
-/p:ask "no sé qué hacer ahora"
-```
-
-**Output:**
-```
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-🎯 Entiendo: necesitas orientación sobre qué hacer
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-📊 TU CONTEXTO:
-  • Estado actual: sin tarea activa ✅
-  • Tareas en cola: 5 tareas esperando
-  • Último ship: hace 4 días ⚠️
-
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-💡 TE RECOMIENDO:
-
-1. VER TU COLA DE TAREAS
-   /p:next
-   → Muestra las 5 tareas top priority
-   → Ya están priorizadas y listas para trabajar
-
-2. EMPEZAR LA TAREA #1
-   /p:build 1
-   → Empieza automáticamente la tarea top priority
-   → Asigna agente apropiado
-   → Tracking de tiempo
-
-3. VER SUGERENCIAS INTELIGENTES
-   /p:suggest
-   → Analiza tu contexto completo
-   → Recomienda basado en:
-     - Tiempo sin ship (4 días)
-     - Velocidad reciente
-     - Urgencias
-
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-🎯 MI SUGERENCIA ESPECÍFICA:
-
-Han pasado 4 días sin ship → Es momento de completar algo
-
-Opción A: Trabaja en tareas de cola
-   /p:next → /p:build 1 → /p:done → /p:ship
-
-Opción B: Crea algo nuevo rápido
-   /p:feature "quick win" → /p:build 1 → ship today
-
-━━━━━━━━━━━━━━━━━━━━━━━━━━━
-
-¿Qué prefieres?
-1. Ver qué hay en cola (/p:next)
-2. Sugerencias personalizadas (/p:suggest)
-3. Contarme qué tienes en mente
-```
-
-## Key Principles
-
-1. **Conversational**: Natural language, not technical
-2. **Educational**: Explain WHY each command
-3. **Contextual**: Always check project state first
-4. **Interactive**: Ask confirmation, offer choices
-5. **No auto-execution**: Show the path, let user decide
-6. **Bilingual**: Support English and Spanish naturally
-7. **Examples**: Show similar use cases when relevant
-
 ## Validation
 
-- **Optional**: Can run without project initialized
-- **If not initialized**: Suggest `/p:init` first
+- **Optional**: Can run without project initialized (suggests `/p:init` if needed)
 - **Read-only**: Never modifies files, only recommends
 
 ## Edge Cases
 
-### No prjct project
-```
-⚠️  No prjct project here
-
-First, initialize:
-/p:init                    # For existing project
-/p:init "your idea"        # For new project
-
-Then I can help you!
-```
-
-### Very vague request
-```
-🤔 Puedes darme más detalles?
-
-Qué quieres hacer:
-• Agregar nueva funcionalidad?
-• Arreglar algo que no funciona?
-• Mejorar el código existente?
-• No estás seguro?
-
-Cuéntame más y te ayudo a encontrar el comando correcto!
-```
-
-## Success Criteria
-
-After using `/p:ask`, user should:
-- ✅ Understand WHAT commands to use
-- ✅ Understand WHY those commands
-- ✅ Understand the SEQUENCE of actions
-- ✅ Feel confident to proceed
-- ✅ Learn the system while using it
+**No prjct project**: Suggest `/p:init` or `/p:init "idea"` first
+**Vague request**: Ask for more details - feature? bug? optimization? not sure?
+**Active task exists**: Offer options - finish first (recommended) or add to queue

package/templates/commands/bug.md

@@ -1,79 +1,20 @@
 ---
 allowed-tools: [Read, Write]
-description: 'Report and track bugs with auto-prioritization'
+description: 'Report bug with auto-priority'
 ---
 
 # /p:bug
 
-## Usage
-
-```
-/p:bug "<description>"
-```
-
-## What It Does
-
-1. **Severity detection**: Analyzes bug description for severity indicators
-2. **Priority placement**: Adds to `core/next.md` based on severity
-3. **Memory tracking**: Logs bug report to `memory/context.jsonl`
-4. **Quick resolution**: Suggests immediate actions
-
-## Severity Detection
-
-Auto-detects severity from keywords:
-
-- **Critical**: crash, down, broken, not working, production, urgent
-- **High**: error, fail, issue, problem, major
-- **Medium**: bug, incorrect, wrong, should
-- **Low**: minor, small, cosmetic, typo
+## Severity Keywords
+- **Critical**: crash, down, broken, production
+- **High**: error, fail, issue
+- **Medium**: bug, incorrect
+- **Low**: minor, typo
 
 ## Flow
+1. Detect severity → Format: `🐛 [LEVEL] {desc}`
+2. Insert: top if critical/high, bottom if med/low
+3. Update `core/next.md` + log
 
-1. Detect severity from description
-2. Read `core/next.md`
-3. Format bug entry: `🐛 [SEVERITY] description`
-4. Insert at top if critical/high, bottom if medium/low
-5. Write to `core/next.md`
-6. Log to `memory/context.jsonl`
-
-## Response Format
-
-```
-🐛 Bug reported: {severity}
-
-{description}
-
-Priority: {critical/high/medium/low}
-Added to: core/next.md (position {top/bottom})
-
-Quick actions:
-• "start fixing" → /p:now "{description}"
-• "show queue" → /p:next
-
-/p:now | /p:next
-```
-
-## Example
-
-```
-User: p. bug "login button crashes on mobile"
-
-Claude detects:
-- Keywords: "crashes" → CRITICAL severity
-- Action: Add to top of next.md
-- Format: 🐛 [CRITICAL] login button crashes on mobile
-
-Response:
-🐛 Bug reported: CRITICAL
-
-login button crashes on mobile
-
-Priority: critical
-Added to: core/next.md (top priority)
-
-Quick actions:
-• "start fixing" → /p:now "login button crashes on mobile"
-• "show queue" → /p:next
-
-/p:now | /p:next
-```
+## Response
+`🐛 {severity} | Queued #{position} | Start: /p:now`

package/templates/commands/build.md

@@ -1,44 +1,17 @@
 ---
-allowed-tools: [Read, Write]
-description: 'Start task with agent assignment'
+allowed-tools: [Read, Write, GetTimestamp]
+description: 'Start task'
+timestamp-rule: 'GetTimestamp() for id and started'
 ---
 
 # /p:build
 
-## Validation
-
-- Blocks if: `core/now.md` has active task
-- Error: "Complete current task first. Use /p:done"
+## Check
+Block if `core/now.md` active
 
 ## Flow
-
-1. Parse: task desc OR queue # (1-5)
-2. Detect: `TaskSchema.detectAgent(task)` + complexity
-3. Get: GitHub dev from `git config user.name` + remote
-4. Create: task metadata with agent, complexity, estimate
-5. Write: `core/now.md` with frontmatter
-
-## Frontmatter
-
-```yaml
----
-id: task-{timestamp}
-agent: {detected_agent}
-dev: @{github_username}
-complexity: {trivial|simple|moderate|complex|epic}
-estimated: {time_range}
-started: {iso_timestamp}
----
-{task_description}
-```
+1. Parse task/# → Detect agent + complexity
+2. Write now.md: id={GetTimestamp()}, agent, started={GetTimestamp()}
 
 ## Response
-
-```
-🎯 {task}
-{agent_icon} {agent_name}
-Est: {time_estimate}
-Complexity: {level}
-
-/p:done when finished
-```
+`🎯 {task} | {agent} | Est: {time} | Done: /p:done`

package/templates/commands/cleanup.md

@@ -1,42 +1,19 @@
 ---
 allowed-tools: [Read, Edit, Bash]
-description: 'Code cleanup and optimization'
+description: 'Code cleanup'
 ---
 
 # /p:cleanup
 
-## Usage
-
-```
-/p:cleanup [--type code|imports|files|deps|all] [--dry-run]
-```
-
-## Flow
-
-1. Parse: type + mode
-2. Backup: create if needed
-3. Execute: cleanup by type
-4. Validate: syntax check
-5. Log: `memory/cleanup-log.jsonl`
-
 ## Types
+- **code**: Remove logs, dead code
+- **imports**: Clean unused
+- **files**: Remove temp/empty
+- **deps**: Find unused
+- **all**: Everything
 
-- **code**: Remove console.logs, commented code, dead code
-- **imports**: Remove unused, organize imports
-- **files**: Remove temp files, empty files
-- **deps**: Analyze unused dependencies
-- **all**: All cleanup types
+## Flow
+Parse type → Backup → Clean → Validate → Log
 
 ## Response
-
-```
-🧹 Cleanup complete!
-
-📊 Results:
-• Console.logs: {N} removed
-• Dead code: {N} blocks
-• Unused imports: {N}
-• Space freed: {X} MB
-
-/p:test | /p:ship
-```
+`🧹 Cleaned: {N} logs, {N} dead code, {N} imports | Freed: {X}MB`