Code Reviewer (Official Plugin)

Integrates the official Claude Code code-review plugin.

v2.88 Key Changes (MODEL-AGNOSTIC)

• Model-agnostic: Uses model configured in ~/.claude/settings.json or CLI/env vars
• No flags required: Works with the configured default model
• Flexible: Works with GLM-5, Claude, Minimax, or any configured model
• Settings-driven: Model selection via ANTHROPIC_DEFAULT_*_MODEL env vars

Quick Start

/code-review              # Review current changes
/code-review --comment    # Post review as PR comment

Architecture

4 parallel agents with confidence scoring (≥80):

1. Agent #1: CLAUDE.md compliance
2. Agent #2: CLAUDE.md compliance (redundancy)
3. Agent #3: Bug detection (changes only)
4. Agent #4: Git blame/history analysis

Features

• Multi-agent parallel review
• Confidence-based filtering
• Auto-skip closed/draft/trivial PRs
• Direct GitHub code links

Integration

Official plugin: ~/.claude-sneakpeek/zai/config/plugins/cache/anthropics/code-review/

Author: Boris Cherny ([email protected]) Version: 1.0.0

Agent Teams Integration (v2.88)

Optimal Scenario: Pure Custom Subagents

This skill uses Pure Custom Subagents (no Agent Teams) for focused code review operations.

Why Scenario B for Code Review

• Native multi-agent: Already uses 4 parallel agents via official plugin
• Read-only operation: Review should not modify code (tool restrictions critical)
• No coordination needed: Analysis is independent per file/PR
• Faster execution: Direct spawn avoids team creation overhead

Configuration

1. No TeamCreate: Official plugin already handles parallelism
2. Direct Task: Spawn ralph-reviewer for additional analysis if needed
3. Tool Restrictions: Review agents restricted to Read/Grep/Glob (no Write/Edit)
4. Model Inheritance: Use model configured in settings.json

Workflow Pattern

# Official plugin runs 4 parallel agents natively
/code-review → Plugin spawns 4 agents with confidence scoring

# For extended analysis (optional):
Task(subagent_type="ralph-reviewer", prompt="Deep analysis of specific file...")
  → Agent executes with restricted tools
  → Returns detailed findings
  → Complete

Tool Restrictions for Safety

When NOT to Use Agent Teams

• Official plugin already provides parallel review
• Read-only analysis doesn't need team coordination
• Tool restrictions prevent accidental changes more effectively than teams
• When review is a one-shot operation (not multi-phase workflow)

Action Reporting (v2.93.0)

Esta skill genera reportes automáticos completos para trazabilidad:

Reporte Automático

Cuando esta skill completa, se genera automáticamente:

1. En la conversación de Claude: Resultados visibles
2. En el repositorio: docs/actions/code-reviewer/{timestamp}.md
3. Metadatos JSON: .claude/metadata/actions/code-reviewer/{timestamp}.json

Contenido del Reporte

Cada reporte incluye:

• ✅ Summary: Descripción de la tarea ejecutada
• ✅ Execution Details: Duración, iteraciones, archivos modificados
• ✅ Results: Errores encontrados, recomendaciones
• ✅ Next Steps: Próximas acciones sugeridas

Ver Reportes Anteriores

# Listar todos los reportes de esta skill
ls -lt docs/actions/code-reviewer/

# Ver el reporte más reciente
cat $(ls -t docs/actions/code-reviewer/*.md | head -1)

# Buscar reportes fallidos
grep -l "Status: FAILED" docs/actions/code-reviewer/*.md

Generación Manual (Opcional)

source .claude/lib/action-report-lib.sh
start_action_report "code-reviewer" "Task description"
# ... ejecución ...
complete_action_report "success" "Summary" "Recommendations"

Referencias del Sistema

• Action Reports System - Documentación completa
• action-report-lib.sh - Librería helper
• action-report-generator.sh - Generador