Claude Code Plugins

Community-maintained marketplace

Feedback

pal-codereview

@estiens/metacognitive_chat
0
0

Systematic code review covering quality, security, performance, and architecture using PAL MCP. Use for pull request reviews, code audits, or pre-commit validation. Triggers on review requests, PR reviews, or code quality checks.

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

name pal-codereview
description Systematic code review covering quality, security, performance, and architecture using PAL MCP. Use for pull request reviews, code audits, or pre-commit validation. Triggers on review requests, PR reviews, or code quality checks.

PAL Code Review

Structured code review with expert validation across quality, security, performance, and architecture.

When to Use

  • Pull request reviews
  • Pre-merge code audits
  • Security-focused reviews
  • Performance reviews
  • Code quality assessments

Quick Start

# Start a code review
result = mcp__pal__codereview(
    step="Reviewing authentication changes for security and quality",
    step_number=1,
    total_steps=2,
    next_step_required=True,
    findings="Beginning comprehensive review",
    review_type="full",
    relevant_files=[
        "/app/auth/login.py",
        "/app/auth/tokens.py"
    ],
    confidence="exploring"
)

Review Types

Type Focus
full Comprehensive: quality, security, performance, architecture
security Security vulnerabilities and auth issues
performance Bottlenecks and optimization
quick Fast pass for obvious issues

Validation Types

Type Description
external Uses expert model for validation (default)
internal Local-only review, no external validation

Required Parameters

Parameter Type Description
step string Review narrative
step_number int Current step
total_steps int Total steps (usually 2 for external)
next_step_required bool More review needed?
findings string Review findings

Optional Parameters

Parameter Type Description
review_type enum full/security/performance/quick
review_validation_type enum external/internal
severity_filter enum critical/high/medium/low/all
focus_on string Specific emphasis (e.g., "threading")
standards string Coding standards to enforce
relevant_files list Files under review
issues_found list Issues with severity
continuation_id string Continue session

Issue Severity Levels

issues_found=[
    {"severity": "critical", "description": "SQL injection vulnerability"},
    {"severity": "high", "description": "Missing authentication check"},
    {"severity": "medium", "description": "Inefficient database query"},
    {"severity": "low", "description": "Inconsistent naming convention"}
]

Example: Security Review

mcp__pal__codereview(
    step="Security-focused review of payment processing code",
    step_number=1,
    total_steps=2,
    next_step_required=True,
    findings="Scanning for injection, auth bypass, data exposure",
    review_type="security",
    severity_filter="medium",  # Report medium and above
    relevant_files=[
        "/app/payments/processor.py",
        "/app/payments/validation.py"
    ],
    focus_on="input validation, PCI compliance",
    confidence="exploring"
)

Review Checklist

Quality

  • Code readability and clarity
  • Appropriate abstractions
  • Error handling
  • Test coverage

Security

  • Input validation
  • Authentication/authorization
  • Secrets management
  • Injection prevention

Performance

  • Database query efficiency
  • Caching strategy
  • Algorithm complexity
  • Resource cleanup

Architecture

  • Separation of concerns
  • Dependency management
  • API design
  • Scalability considerations

Best Practices

  1. Cover all aspects - Quality, security, performance, architecture
  2. Prioritize findings - Critical issues first
  3. Be constructive - Suggest improvements, not just problems
  4. Reference code - Use file paths and line numbers
  5. Consider context - Time pressure, team experience