Project Reviewer

Automated project-wide code review tool that scans Python projects and identifies files needing refactoring based on Clean Code principles.

Quick Start

When triggered by /review-project <directory_path>, this skill will:

1. •Recursively scan the directory for all .py files
2. •Apply Clean Code checks to each file
3. •Generate an integrated report sorted by file path

Required MCP Tools

This skill requires the following MCP tools to be available:

• •list_directory - For directory traversal and file discovery
• •read_file - For reading Python file contents

If these tools are not available, the skill will fail with an error message.

Workflow

Step 1: Directory Scanning

Use list_directory tool recursively to discover all .py files:

1. •Start with the provided directory path
2. •List all entries in the directory
3. •For each entry:
   • •If it's a .py file, add to the review queue
   • •If it's a subdirectory, recursively scan it
4. •Skip common directories: __pycache__, .git, .venv, venv, node_modules, .pytest_cache

Step 2: File Review

For each Python file found, use read_file tool and apply Clean Code checks:

Variable Naming Check:

• •Single letter names (except i, j, k in loops)
• •Unclear abbreviations (d, tmp, val, etc.)
• •Non-descriptive names (data, info, obj)
• •Names not revealing intent (flag, check)

Function Length Check:

• •Target: Maximum 20 lines (excluding blank lines and docstrings)
• •Flag functions exceeding this limit

Redundant Comments Check:

• •Comments repeating what code clearly states
• •Comments explaining obvious operations
• •Comments that could be replaced by better naming

See clean-code-rules.md for detailed checking criteria.

Step 3: Generate Integrated Report

Produce a comprehensive report with:

1. •

   Executive Summary

   • •Total files scanned
   • •Total issues found
   • •Top 3 files needing refactoring (by issue count)
2. •

   Detailed Results (sorted by file path)

   • •File path
   • •Issue counts by category (naming, length, comments)
   • •Specific issues with line numbers
3. •

   Recommendations

   • •Priority ranking for refactoring
   • •Common patterns observed across files

Report Format

=== Project Code Review Report ===

Executive Summary:
- Files Scanned: 15
- Total Issues: 47
- Top 3 Files Needing Refactoring:
  1. src/utils/helper.py (12 issues)
  2. src/main.py (9 issues)
  3. tests/test_api.py (8 issues)

=== Detailed Results ===

File: src/main.py (9 issues)
├─ Variable Naming (4):
│  ├─ Line 12: 'd' -> Use descriptive name
│  ├─ Line 25: 'tmp' -> Rename to indicate purpose
│  ├─ Line 40: 'x' -> Use meaningful name
│  └─ Line 55: 'flag' -> Use descriptive boolean name
├─ Function Length (2):
│  ├─ process_data() (lines 30-55): 26 lines
│  └─ handle_request() (lines 60-85): 26 lines
└─ Redundant Comments (3):
   ├─ Line 10: "// Set name" - Remove
   ├─ Line 35: "// Loop through items" - Remove
   └─ Line 70: "// Return result" - Remove

File: src/utils/helper.py (12 issues)
[... similar format ...]

=== Recommendations ===

Priority 1 - Critical (>10 issues):
- src/utils/helper.py: Heavy refactoring needed

Priority 2 - High (6-10 issues):
- src/main.py: Focus on variable naming and function splitting
- tests/test_api.py: Improve test function names

Priority 3 - Medium (3-5 issues):
- src/api/routes.py: Minor improvements needed

Priority 4 - Low (1-2 issues):
- src/config.py: Minimal changes required

Common Patterns:
- Frequent use of single-letter variables across 8 files
- Long functions (>20 lines) found in 5 files
- Redundant comments in 10 files

Error Handling

Handle common errors gracefully:

• •Directory not found: Display clear error message with the attempted path
• •No .py files found: Report that the directory contains no Python files
• •File read errors: Skip the file and note it in the report
• •MCP tools unavailable: Fail fast with clear message about missing tools

Performance Considerations

For large projects (>100 files):

• •Process files in batches
• •Provide progress updates
• •Consider offering to focus on specific subdirectories

Integration with clean-code-reviewer

This skill uses the same checking logic as clean-code-reviewer but applies it across multiple files. The rules are consistent:

• •Python: 20 line function limit
• •Same variable naming conventions
• •Same comment redundancy criteria

For language-specific Python rules, refer to the clean-code-reviewer skill's Python guidelines.