Package Exports

@bobmatnyc/ai-code-review
@bobmatnyc/ai-code-review/dist/index.js

This package does not declare an exports field, so the exports above have been automatically detected and optimized by JSPM instead. If any package subpath is missing, it is recommended to post an issue to the original package (@bobmatnyc/ai-code-review) to support the "exports" field. If that is not possible, create a JSPM override to customize the exports field for this package.

Readme

AI Code Review v4.4.5

A TypeScript-based tool for automated code reviews using Google's Gemini AI models, Anthropic Claude models (including Claude 4), OpenAI models, and OpenRouter API with LangChain integration for enhanced prompt management.

What's New in v4.4.5

Documentation Improvements in v4.4.5

Comprehensive Documentation Review: Updated all documentation to ensure consistency with current codebase
Version Consistency: Updated all version references across README.md, INSTALL.md, and CLAUDE.md
Enhanced CLAUDE.md: Added comprehensive project context, development guidelines, and troubleshooting
Review Types Documentation: Updated to include all 15 current review types including new ai-integration, cloud-native, and developer-experience types
CLI Documentation: Fixed command examples and ensured all documented features match current implementation
Model References: Updated to current recommended models (gemini-2.5-pro, claude-4-sonnet)

What's New in v4.4.4

New Features in v4.4.4

Unified Client System: Revolutionary architecture for API client management
- BaseApiClient interface for consistent client behavior across all providers
- UnifiedClientFactory for streamlined client creation and configuration
- Enhanced OpenAIApiClient implementation with improved error handling
- Better separation of concerns and improved maintainability
Build Number Tracking: Comprehensive build tracking and versioning system
- Automatic build number incrementation during builds
- Build metadata tracking with timestamps for better release management
- Enhanced version reporting with detailed build information
Enhanced Consolidation Service: Improved multi-pass review consolidation
- Dedicated ConsolidationService for better separation of concerns
- Enhanced error handling and fallback mechanisms for robust operation
- Improved AI-powered consolidation with better prompt handling
Extended Prompt Templates: Expanded template library for specialized reviews
- AI Integration Review template for AI-assisted development analysis
- Cloud Native Review template for cloud architecture assessment
- Developer Experience Review template for DX evaluation
- Java and Rust language-specific best practices templates
Prompt Schema Validation: JSON schema validation for prompt frontmatter
- Structured validation for prompt metadata consistency
- Consistent prompt template format enforcement
- Enhanced prompt development workflow with validation

Improvements in v4.4.4

Documentation Organization: Comprehensive documentation restructuring
- Moved all documentation files to dedicated docs/ directory
- Enhanced documentation for unified client system
- Improved build number tracking documentation
- Better organization of migration guides and architectural docs
Code Quality: Enhanced testing and validation infrastructure
- New unit tests for consolidation bug fixes
- Improved test coverage for unified client system
- Enhanced validation scripts for prompt templates
Project Structure: Cleaner codebase organization
- Replaced old task management system with new tickets/ structure
- Removed temporary debug files and test directories
- Better separation of concerns in core services

Fixed in v4.4.4

Consolidation Bug: Fixed critical issue where consolidation instructions were treated as source code
Token Analysis: Improved token counting and analysis accuracy
Client Factory: Enhanced client creation and initialization logic

What's New in v4.3.0

Key Features in v4.3.0

Prompt Template Manager: Handlebars integration for dynamic prompt templates with conditional logic and reusable components
Extract Patterns Review Type: Phase 2 implementation with pattern validation and structured output
Output Directory Override: New --output-dir flag for custom review output locations
YAML Configuration Support: .ai-code-review.yaml configuration files with automatic discovery
Enhanced Testing: Achieved 96.8% test pass rate with comprehensive framework implementation

What's New in v4.2.0

Key Features in v4.2.0

Evaluation Review Type: Comprehensive code review with developer skill assessment (Beginner to Expert levels)
Golang Language Support: Complete Go/Golang project support with idiomatic patterns and best practices
Enhanced Grading System: Academic-style A+ to F grading for code quality assessment

What's New in v4.1.x

Key Changes in v4.1.0

Deprecated --individual flag (single-file behavior is automatic based on target)
Repurposed -i as alias for --interactive mode for better user experience
Simplified codebase by removing redundant individual file review logic

What's New in v4.0.x

Major Features in v4.0.0

🧠 AI-Guided Semantic Chunking: Revolutionary TreeSitter-based code analysis with 95%+ token reduction
🎯 Intelligent Strategies: 5 chunking strategies optimized for different review types and code structures
⚡ Performance: Solved context limit problems, reducing typical 196K+ token reviews to ~4K tokens

Fixed in v4.0.2

🔧 Complete TypeScript Error Resolution: Fixed all 15 TypeScript compilation errors
🧹 Code Quality Improvements: Achieved zero ESLint errors with proper type safety
✅ Build Pipeline Stability: Full compilation success with clean CI/CD pipeline

Version History

For detailed version history and release notes, see CHANGELOG.md.

Recent Releases

v4.4.5 (2025-08-17): Documentation consistency improvements, enhanced CLAUDE.md, updated review types documentation
v4.4.4 (2025-08-16): Unified Client System, Build Number Tracking, Enhanced Consolidation Service
v4.3.0 (2025-06-30): Prompt Template Manager, Extract Patterns Review Type, YAML Configuration Support
v4.2.0 (2025-06-24): Evaluation Review Type, Golang Language Support, Enhanced Grading System
v4.1.0 (2025-06-08): CLI improvements, deprecated --individual flag, enhanced user experience
v4.0.0 (2025-06-04): AI-Guided Semantic Chunking, 95%+ token reduction, intelligent strategies

v3.2.15 (2025-06-01)

Re-added O3 Models: Re-added support for OpenAI o3 and o3-mini reasoning models
Model Configuration: Properly configured o3 models with max_completion_tokens parameter
Better Understanding: The issue may have been with multi-pass consolidation, not the models themselves

v3.2.14 (2025-06-01)

Multi-Pass Consolidation Fix: Fixed issue where non-OpenAI providers were reviewing consolidation content as source code
Gemini Consolidation: Improved consolidation logic for Gemini and other providers to properly merge multi-pass results
Better Error Handling: Added fallback consolidation when AI consolidation fails

v3.2.13 (2025-06-01)

Environment Override Fix: Fixed project .env.local not overriding tool's environment variables
Project Priority: Ensured project-level .env.local files properly override tool installation settings
Debug Logging: Added detailed environment variable tracing to diagnose loading issues

v3.2.12 (2025-06-01)

CLI Fix: Fixed -m option not overriding environment variable
Environment Loading: Improved priority order - CLI > project .env files > system env

v3.2.11 (2025-06-01)

OpenAI Model Removal: Removed o3 and o3-mini models per issue #44 (generating generic advice instead of analyzing code)
Test Updates: Fixed model counts in tests after removal

v3.2.10 (2025-05-29)

Documentation Updates: Comprehensive README updates with full version history
Build Process: Enhanced publishing checklist

v3.2.9 (2025-05-28)

Claude 4 Model Support: Added Anthropic's latest Claude 4 Opus and Sonnet models with 200K context windows
Fixed Provider Display: Resolved issue where Anthropic models showed "Gemini" as provider
Improved Cost Estimation: Fixed token/cost calculation accuracy and display clarity

v3.2.8 (2025-05-21)

Critical Version Fix: Fixed CLI reporting incorrect version (3.0.3 instead of actual version)
Version Management: Added comprehensive version management documentation

v3.2.7 (2025-05-22)

OpenAI Model Cleanup: Removed non-working reasoning models (o3, o4-mini, o4-mini-high)
API Stability: Cleaned up OpenAI client to only include publicly available models

v3.2.6 (2025-05-22)

New OpenAI Models: Added o3, o4-mini, and o4-mini-high reasoning models (later removed in v3.2.7)

v3.2.5 - v3.2.3 (2025-05-22)

Type Safety Improvements: Comprehensive TypeScript strictness improvements
Security Enhancements: Replaced any types with proper interfaces
CI/CD Improvements: Fixed all lint errors and test failures
Smart File Selection: Added .eslintignore and tsconfig.json support

v3.2.2 - v3.2.0 (2025-05-21)

JSON Configuration: Added support for external config files
Framework Templates: Added specialized templates for Angular, Vue, Django, Laravel
CI Integration: Added npm run ci:local for pre-push validation
Handlebars Templates: Migrated prompt system to Handlebars for better maintainability

v3.1.0 (2025-05-20)

Enhanced Prompts: Improved prompt templates with YAML metadata
Dependency Analysis: Added dependency-cruiser and ts-prune integration
Multi-Language Support: Added specialized prompts for Ruby/Rails, Python, PHP, Golang

What's New in v3.0.2

Improvements in v3.0.2

Better Default Behavior: Automatically uses current directory when no target path is provided
Clearer Output: Fixed issues with undefined review titles and model information
Improved Logging: Added informative logs when using default paths
Fixed Multi-Pass Reviews: Resolved JSON parsing errors in multi-pass reviews with code blocks
Enhanced Regex: Improved patterns for extracting JSON from various code block formats
More Robust: Enhanced error handling for missing data in review results

What's New in v3.0.1

Added in v3.0.1

Enhanced Review Structure: Added grade section to OpenAI reviews
Consistent Formatting: All model providers now use consistent review format with grades
Detailed Grading: Added granular grade categories covering functionality, code quality, documentation, etc.

What's New in v3.0.0

Major Features in v3.0.0

Smart Multi-Pass Reviews: Added intelligent confirmation for multi-pass reviews showing token usage, estimated passes, and cost
Improved CLI Experience: Enhanced command-line interface with better organization and clearer documentation
Enhanced Model Support: Updated support for the latest Gemini 2.5 Pro models and API specifications
Comprehensive Test Suite: Added extensive test coverage for command-line options

Added in v3.0.0

Multi-Pass Confirmation: Added interactive confirmation before proceeding with multi-pass reviews
--no-confirm Flag: Skip confirmation for multi-pass reviews when automated processing is needed
Cost Estimation: Show detailed token usage and cost estimates before large multi-pass operations
Improved Documentation: Better organized and more detailed command documentation

What's New in v2.2.0

Added in v2.2.0

Ruby/Rails Support: Complete support for Ruby and Ruby on Rails projects
- Specialized Ruby/Rails prompt templates for all review types
- Rails-specific best practices and conventions
- Ruby project type detection for framework-aware analysis
- Support for Ruby-specific file extensions and directory structures
- Comprehensive Rails security review capabilities
- Performance analysis focused on ActiveRecord and Rails optimizations
Enhanced Reviews: Added file list and comprehensive dependency security analysis
- Automatically includes a list of all files analyzed in architectural reviews
- Performs security vulnerability analysis on package dependencies
- Detects technology stack (Next.js, Laravel, Django, Rails, etc.) for framework-specific analysis
- Identifies security issues with severity ratings, CVE IDs, and recommended updates
- Improves visibility into what was examined during the review
- See ENHANCED_REVIEW.md, SECURITY_ANALYSIS.md, and STACK_DETECTION.md for usage details

What's New in v2.1.1

Fixes and Improvements in v2.1.1

Enhanced Metadata Headers: Added comprehensive metadata headers to each review with model details, token usage, cost, tool version, and command options used to generate the review
- Markdown reviews now include a detailed metadata table
- JSON reviews contain a structured metadata object
- Both formats capture timestamp, model information, token usage, and command details
Ruby/Rails Support: Added specialized Ruby/Rails review capabilities
- Language-specific prompt templates for more accurate Ruby/Rails reviews
- Ruby/Rails project type detection for framework-aware analysis
- Support for Ruby-specific file extensions (.rb, .rake, .gemspec, .ru, .erb)
- Rails-specific directory exclusions for better performance
Version Display: Added version display at startup for easier version identification
Enhanced Python Support: Improved Python file detection and enhanced file globbing patterns
Better Path Display: Fixed issue where review output did not show full directory path
Improved Installation: Added global installation utilities to prevent conflicts
Troubleshooting: Added ./scripts/fix-global-command.sh script to quickly resolve global installation issues
Fixed Sanitization: Improved content sanitization to properly preserve newlines and tabs

Major Features in v2.1.0

Automatic Language Detection: Automatically detects Ruby, Python, PHP, Golang, TypeScript, and JavaScript projects
Multi-Language Support: Added specialized prompts for Ruby/Rails, Python, PHP, and Golang code review
Dependency Analysis: Added dependency-cruiser integration for architectural reviews
Static Code Analysis: Added ts-prune and eslint integration for unused code detection
Enhanced Prompts: Improved prompt templates with standardized YAML metadata
TypeScript Focus: Added specialized TypeScript architectural review prompt

Major Features in v2.0.0

Stable Release: Version 2.0.0 marks the first stable release with all major features complete
LangChain Integration: Added LangChain for enhanced prompt management, templating, and chain capabilities
New Review Types: Added 'unused-code' review type to identify and suggest removal of dead code
Improved Prompts: Enhanced prompt templates with LangChain, including few-shot learning and structured output
Structured Schemas: Created detailed Zod schemas for all review types to enable more structured and consistent output
Enhanced TypeScript Support: Added TypeScript-specific templates and analyzers for better static analysis
Model Testing: Added new commands to test individual models (model-test) and verify all models on build
- Test specific models with ai-code-review model-test gemini:gemini-2.5-pro
- Test all models from a provider with ai-code-review model-test -p gemini
- Test all available models with ai-code-review model-test --all
- Verify models during build with ai-code-review test-build

Other Improvements

Fixed Unit Tests: Ensured compatibility with the latest dependencies
Improved Jest Configuration: Added proper handling for ESM modules
Removed Prettier Checking: Improved developer experience by removing Prettier from the test process
Fixed ESLint Issues: Added p-limit dependency to fix ESLint errors
Model Listing Feature: Added --models flag to list all supported models with their configuration names
Improved Error Handling: Enhanced error handling and recovery mechanisms
Debug Logging Control: Suppressed DEBUG logging messages in production builds
Performance Optimizations: Improved memory usage and processing speed
Build Verification: Added automatic model testing during build process

What's New in v1.3.2

Fixed API Client Selection: Fixed bug in dynamic imports for API clients
Fixed Anthropic API Version: Updated Anthropic API version to use the correct version

What's New in v1.3.1

Simplified Model Names: Removed version-specific details from model names for better usability
Improved Model Management: Derived model lists from a single source of truth
Structured Output: Added structured JSON output format for code reviews
Enhanced Formatting: Improved formatting of review results with priority-based grouping

What's New in v1.3.0

Structured Output Format: Added structured JSON output for code reviews
JSON Parsing: Added support for parsing JSON responses wrapped in code blocks
New Type Definitions: Added structured review type definitions
Improved Formatting: Added formatStructuredReviewAsMarkdown function

What's New in v1.2.0

Multi-Provider Support: Added support for multiple AI providers (Google, Anthropic, OpenAI, OpenRouter)
Token and Cost Estimation: Implemented comprehensive token and cost estimation for all supported models
Model Listing: Added --listmodels flag to display all available models
Improved Code Organization: Reorganized utility modules to reduce duplication and improve maintainability
Enhanced Documentation: Added detailed JSDoc comments to key functions and classes
Bug Fixes: Fixed various issues including language output problems and failing tests

Overview

This tool analyzes code from specified files or directories in sibling projects and generates structured code evaluations. It leverages Google's Gemini AI models, Anthropic Claude models, and OpenRouter API to provide insightful feedback on code quality, best practices, and potential improvements. With support for multiple AI models, you can choose the best model for your specific needs.

Features

Multiple Review Types: Focus on different aspects of code quality:
- quick-fixes: Fast issue identification and recommendations
- architectural: Deep structural analysis and design patterns
- security: Security vulnerability detection and remediation
- performance: Performance bottleneck identification
- unused-code: Dead code detection and cleanup recommendations
- focused-unused-code: Focused unused code analysis
- code-tracing-unused-code: Code tracing for unused code detection
- consolidated: Consolidated review combining multiple approaches
- best-practices: Best practices analysis and recommendations
- evaluation: Comprehensive code evaluation with developer skill assessment
- extract-patterns: Code pattern analysis and best practice suggestions
- coding-test: Comprehensive coding test evaluation with assignment scoring and AI detection
- ai-integration: AI-assisted development analysis and recommendations
- cloud-native: Cloud architecture assessment and optimization
- developer-experience: Developer experience evaluation and improvements
Interactive Mode: Process review results interactively, implementing fixes based on priority
Automatic Fixes: Automatically implement high priority fixes without manual intervention
Prompt-Based Fixes: Confirm and apply medium and low priority fixes with user input
Directory Support: Review entire directories and their subdirectories in one command
Consolidated Reviews: Generate a single comprehensive review for multiple files
Project Context: Include project documentation in the review context
Multiple AI Models: Support for Google's Gemini models, Anthropic Claude models, and OpenRouter API (Claude, GPT-4, etc.)
Model Listing: List all available models with the --listmodels flag
Token and Cost Estimation: Estimate token usage and cost with the --estimate flag
Customizable: Configure review types, output formats, and prompt templates
GitHub Projects Integration: Sync PROJECT.md content with GitHub Projects for better project management
Memory Optimized: Process large codebases efficiently with optimized memory usage
Error Recovery: Robust error handling with graceful recovery
Code Focus: Analyzes executable code files only, excluding configuration and documentation files

Security Best Practices

⚠️ IMPORTANT: Never commit real API keys to your repository!

Always use environment variables for API keys
Add .env.local to your .gitignore file
Use placeholder values in example files (like your_api_key_here)
Review code before committing to ensure no secrets are included
Consider using secret scanning tools in your CI/CD pipeline

Quick Start

New to AI Code Review? Check out our Quick Start Guide for a 5-minute setup.

Installation

Global Installation

# Using pnpm (recommended)
pnpm add -g @bobmatnyc/ai-code-review

# Using npm (alternative)
npm install -g @bobmatnyc/ai-code-review

Local Installation

# Using pnpm (recommended)
pnpm add -D @bobmatnyc/ai-code-review

# Using npm (alternative)
npm install --save-dev @bobmatnyc/ai-code-review

API Key Setup

Create a .env.local file in your project root with your API keys:

# Required: Model selection
AI_CODE_REVIEW_MODEL=gemini:gemini-2.5-pro
# or
# AI_CODE_REVIEW_MODEL=openrouter:anthropic/claude-3-opus
# or
# AI_CODE_REVIEW_MODEL=anthropic:claude-3-opus

# Required: API key for the selected model type
# For Google Gemini models
AI_CODE_REVIEW_GOOGLE_API_KEY=your_google_api_key_here

# For OpenRouter models (Claude, GPT-4, etc.)
AI_CODE_REVIEW_OPENROUTER_API_KEY=your_openrouter_api_key_here

# For direct Anthropic Claude models
AI_CODE_REVIEW_ANTHROPIC_API_KEY=your_anthropic_api_key_here

You can get API keys from:

Google AI Studio for Gemini models
OpenRouter for access to Claude, GPT-4, and other models
Anthropic for direct access to Claude models

Usage

Command Line

# Global installation
ai-code-review [target] [options]

# Local installation with pnpm
pnpm exec ai-code-review [target] [options]

# Local installation with npm
npx ai-code-review [target] [options]

# Note: The tool only works within the current project
# If no target is specified, it defaults to the current directory (".")

Examples

# Review a single file in the current project
ai-code-review src/index.ts

# Review the current directory (two equivalent commands)
ai-code-review
ai-code-review .

# Review an entire directory with interactive mode
ai-code-review src/utils --interactive

# Perform an architectural review with file list and dependency security analysis
ai-code-review src --type architectural

# Review a Ruby on Rails application with Rails-specific analysis
ai-code-review app --type architectural --language ruby

# Find security vulnerabilities in a Rails application with package security analysis
ai-code-review app --type security
ai-code-review app/controllers --type security --language ruby

# Find unused code that can be safely removed
ai-code-review src --type unused-code

# Use specific review types
ai-code-review src --type focused-unused-code
ai-code-review src --type code-tracing-unused-code
ai-code-review src --type consolidated
ai-code-review src --type best-practices
ai-code-review src --type evaluation
ai-code-review src --type extract-patterns
ai-code-review src --type ai-integration
ai-code-review src --type cloud-native
ai-code-review src --type developer-experience

# Perform coding test evaluation with AI detection
ai-code-review ./candidate-submission --type coding-test --enable-ai-detection

# Advanced coding test with custom threshold and specific analyzers
ai-code-review ./submission --type coding-test \
  --enable-ai-detection \
  --ai-detection-threshold 0.8 \
  --ai-detection-analyzers git,documentation

# Include test files in the review
ai-code-review src --include-tests

# Specify output format (markdown or json)
ai-code-review src/index.ts --output json

# Specify output directory for review results
ai-code-review src --output-dir ./reviews

# Use specific model
ai-code-review src --model anthropic:claude-4-sonnet

# List all available models
ai-code-review --listmodels

# List all supported models with their configuration names
ai-code-review --models

# Test a specific model
ai-code-review model-test gemini:gemini-2.5-pro

# Test all models for a specific provider
ai-code-review model-test --provider anthropic

# Test all available models
ai-code-review model-test --all

# Estimate token usage and cost without performing a review
ai-code-review src/utils --estimate

# Skip confirmation for multi-pass reviews
ai-code-review src/utils --no-confirm

# Use multi-pass review for large codebases
ai-code-review src --multi-pass

# Force single-pass review
ai-code-review src --force-single-pass

# Test API connections before review
ai-code-review src --test-api

# Enable interactive mode
ai-code-review src --interactive

# Check the version of the tool
ai-code-review --version

# Show version information
ai-code-review --show-version

# Run in debug mode for additional logging
ai-code-review src/utils --debug

# Specify programming language and framework
ai-code-review src --language typescript --framework next.js

# Use configuration file
ai-code-review src --config .ai-code-review.yaml

Options

Options:
  # Basic Options
  -t, --type <type>       Type of review (choices: quick-fixes, architectural, security, performance, unused-code, focused-unused-code, code-tracing-unused-code, consolidated, best-practices, evaluation, extract-patterns, coding-test, ai-integration, cloud-native, developer-experience) (default: "quick-fixes")
  -o, --output <format>   Output format (markdown, json) (default: "markdown")
  --output-dir <dir>      Directory to save review output
  -m, --model <model>     Model to use for the review (format: provider:model) (default: "gemini:gemini-2.5-pro")
  -v, --version           Output the current version
  --show-version          Show version information (default: false)
  -h, --help              Display help information
  --debug                 Enable debug logging (default: false)

  # Processing Options
  --include-tests         Include test files in the review (default: false)
  --include-project-docs  Include project documentation in the review context (default: true)
  --include-dependency-analysis Include dependency analysis in the review (default: false)
  --enable-semantic-chunking Enable semantic chunking for intelligent code analysis (default: true)
  
  # Execution Options
  -i, --interactive       Run in interactive mode, processing review results in real-time (default: false)
  --test-api              Test API connections before running the review (default: false)
  --estimate              Estimate token usage and cost without performing the review (default: false)
  
  # Multi-pass Options
  --multi-pass            Use multi-pass review for large codebases (default: false)
  --force-single-pass     Force single-pass review even if token analysis suggests multiple passes are needed (default: false)
  --context-maintenance-factor Context maintenance factor for multi-pass reviews (0-1) (default: 0.15)
  
  # Configuration Options
  --no-confirm            Skip confirmation prompts (default: false)
  --language <language>   Specify the programming language (auto-detected if not specified)
  --framework <framework> Specify the framework (auto-detected if not specified)
  --config <path>         Path to JSON configuration file
  
  # Model and Information Options
  --listmodels            List available models based on configured API keys (default: false)
  --models                List all supported models and their configuration names (default: false)
  
  # API Key Options
  --google-api-key <key>  Google API key for Gemini models
  --openrouter-api-key <key> OpenRouter API key
  --anthropic-api-key <key> Anthropic API key for Claude models
  --openai-api-key <key>  OpenAI API key for GPT models
  
  # AI Detection Options (for coding-test type)
  --enable-ai-detection   Enable AI-generated code detection (default: false)
  --ai-detection-threshold <threshold>  AI detection confidence threshold 0.0-1.0 (default: 0.7)
  --ai-detection-analyzers <analyzers>  Comma-separated list of analyzers: git,documentation,structural,statistical,linguistic (default: "git,documentation")
  --ai-detection-include-in-report      Include AI detection results in review report (default: true)
  --ai-detection-fail-on-detection     Automatically fail evaluation if AI-generated code detected (default: false)

Model Testing Options

Command: model-test
Description: Test AI models to verify API keys and model availability

Arguments:
  [provider:model]        Provider and model to test (e.g. gemini:gemini-2.5-pro, anthropic:claude-3-opus)

Options:
  --all                   Test all available models
  -p, --provider <provider> Test all models for a specific provider
  -h, --help              Display help information

Command: test-build
Description: Test the build by running a simple command

Options:
  -h, --help              Display help information

Output

Review results are stored in the ai-code-review-docs/ directory. For consolidated reviews, the output follows this naming pattern:

ai-code-review-docs/[ai-model]-[review-type]-[file-or-directory-name]-[date].md

For example:

ai-code-review-docs/openai-gpt-4o-quick-fixes-review-src-2024-04-06.md
ai-code-review-docs/gemini-1.5-pro-architectural-review-src-utils-2024-04-06.md

When reviewing a single file, the output filename will include the specific file name:

ai-code-review-docs/[ai-model]-[review-type]-[file-name]-[date].md

Review Output Format

Each review includes comprehensive metadata in a standardized format:

Markdown Format

# Code Review: /path/to/reviewed/file

> **Review Type**: quick-fixes
> **Generated**: April 18, 2025, 3:45:27 PM
> **Reviewed**: /path/to/reviewed/file

---

## Metadata
| Property | Value |
|----------|-------|
| Review Type | quick-fixes |
| Generated At | April 18, 2025, 3:45:27 PM EDT |
| Model Provider | OpenAI |
| Model Name | gpt-4o |
| Input Tokens | 12,345 |
| Output Tokens | 2,456 |
| Total Tokens | 14,801 |
| Estimated Cost | $0.0295 |
| Tool Version | 2.1.1 |
| Command Options | --type=quick-fixes --output=markdown |

[Review content here...]

---

*Generated by Code Review Tool using OpenAI (gpt-4o)*

JSON Format

When using --output json, the review includes a structured metadata object:

{
  "filePath": "/path/to/reviewed/file",
  "reviewType": "quick-fixes",
  "content": "...",
  "timestamp": "2025-04-18T15:45:27.123Z",
  "modelUsed": "openai:gpt-4o",
  "toolVersion": "2.1.1",
  "commandOptions": "--type=quick-fixes --output=json",
  "meta": {
    "model": {
      "provider": "OpenAI",
      "name": "gpt-4o",
      "fullName": "OpenAI (gpt-4o)"
    },
    "review": {
      "type": "quick-fixes",
      "path": "/path/to/reviewed/file",
      "generatedAt": "2025-04-18T15:45:27.123Z",
      "formattedDate": "April 18, 2025, 3:45:27 PM EDT"
    },
    "cost": {
      "inputTokens": 12345,
      "outputTokens": 2456,
      "totalTokens": 14801,
      "estimatedCost": 0.0295,
      "formattedCost": "$0.0295"
    },
    "tool": {
      "version": "2.1.1",
      "commandOptions": "--type=quick-fixes --output=json"
    }
  }
}

This enhanced metadata makes it easier to track and analyze reviews, understand the context in which they were generated, and reproduce the same review conditions if needed.

Configuration

Customizing Prompts

You can customize the review process in several ways:

Prompt Templates

The tool comes with built-in prompt templates in the prompts/templates/ directory:

quick-fixes-review.md - For quick fixes reviews
security-review.md - For security reviews
architectural-review.md - For architectural reviews
performance-review.md - For performance reviews

Custom Prompt Templates

You can create your own prompt templates and use them with the --prompt flag. Custom templates should include metadata in YAML format at the top of the file:

---
name: Custom Security Review
description: A custom prompt template for security-focused code reviews
version: 1.0.0
author: Your Name
reviewType: security
language: typescript
tags: security, custom
---

# Security Code Review

Please review the following code for security vulnerabilities:

{{LANGUAGE_INSTRUCTIONS}}

## Output Format

Please provide your findings in the following format:

1. **Vulnerability**: Description of the vulnerability
2. **Severity**: High/Medium/Low
3. **Location**: File and line number
4. **Recommendation**: How to fix the issue

{{SCHEMA_INSTRUCTIONS}}

Prompt Fragments

You can inject custom fragments into the prompt with the --prompt-fragment flag:

ai-code-review src/index.ts --prompt-fragment "Focus on performance issues"

You can also specify the position of the fragment with the --prompt-fragment-position flag (start, middle, or end):

ai-code-review src/index.ts --prompt-fragment "Focus on security issues" --prompt-fragment-position start

Model-Specific Strategies

You can use model-specific prompt strategies with the --prompt-strategy flag:

ai-code-review src/index.ts --prompt-strategy anthropic

Available strategies:

anthropic - Optimized for Claude models
gemini - Optimized for Gemini models
openai - Optimized for GPT models
langchain - Uses LangChain for enhanced prompt management

The LangChain strategy is particularly useful for complex reviews that benefit from:

Structured output with Zod schemas
Few-shot learning with examples
Chain-based reasoning

# Use LangChain for an unused code review
ai-code-review src/utils --type unused-code --prompt-strategy langchain

# Use LangChain for quick fixes with enhanced prompts
ai-code-review src/components --type quick-fixes --prompt-strategy langchain

Separate Writer Model for Consolidation

For multi-pass reviews and report consolidation, you can specify a separate model for the final report generation. This is useful for optimizing costs by using a more powerful model for code analysis and a cheaper/faster model for report writing.

Using Writer Model

There are two ways to specify a writer model:

Via Command Line:

# Use Claude Opus for analysis, Haiku for report writing
ai-code-review . --model anthropic:claude-3-opus --writer-model anthropic:claude-3-haiku

# Use GPT-4 for analysis, GPT-4o-mini for consolidation
ai-code-review . --model openai:gpt-4 --writer-model openai:gpt-4o-mini

Via Environment Variable:

# In your .env.local file
AI_CODE_REVIEW_MODEL=anthropic:claude-3-opus
AI_CODE_REVIEW_WRITER_MODEL=anthropic:claude-3-haiku

Benefits

Cost Optimization: Report consolidation primarily involves summarization and formatting, not deep code analysis. A cheaper model can handle this effectively with 10-20x cost savings.
Performance: Faster models can speed up the consolidation phase without sacrificing quality.
Flexibility: Use the best model for each task - powerful models for understanding code, efficient models for writing reports.

How It Works

The primary model (specified by --model or AI_CODE_REVIEW_MODEL) performs the code analysis
For multi-pass reviews, each pass uses the primary model
The writer model (specified by --writer-model or AI_CODE_REVIEW_WRITER_MODEL) consolidates the results into the final report
If no writer model is specified, the primary model is used for all tasks

GitHub Projects Integration

You can sync GitHub projects to local directory for better project management:

# Sync GitHub projects to local directory
ai-code-review sync-github-projects --token <token> --org <org> --output-dir ./projects

Available Subcommands

ai-code-review code-review [target] - Run a code review on the specified target (default)
ai-code-review test-model - Test the configured model with a simple prompt
ai-code-review test-build - Test the build by running a simple command
ai-code-review sync-github-projects - Sync GitHub projects to local directory
ai-code-review generate-config - Generate a sample configuration file

sync-github-projects Options

--token <token> - GitHub token [required]
--org <org> - GitHub organization [required]
--output-dir <dir> - Output directory (default: "./github-projects")

generate-config Options

-o, --output <path> - Output file path for the configuration
-f, --format <format> - Configuration file format (choices: yaml, json) (default: "yaml")
--force - Overwrite existing configuration file (default: false)

Generate a sample configuration file:

# Generate YAML configuration
ai-code-review generate-config --output .ai-code-review.yaml --format yaml

# Generate JSON configuration
ai-code-review generate-config --output .ai-code-review.json --format json

Environment Variables

All environment variables use the AI_CODE_REVIEW_ prefix:

# API Keys
AI_CODE_REVIEW_GOOGLE_API_KEY=your_google_api_key_here
AI_CODE_REVIEW_ANTHROPIC_API_KEY=your_anthropic_api_key_here
AI_CODE_REVIEW_OPENROUTER_API_KEY=your_openrouter_api_key_here
AI_CODE_REVIEW_OPENAI_API_KEY=your_openai_api_key_here

# Model Configuration
AI_CODE_REVIEW_MODEL=gemini:gemini-2.5-pro
AI_CODE_REVIEW_WRITER_MODEL=openai:gpt-4o-mini

# Debug Options
AI_CODE_REVIEW_LOG_LEVEL=info

Global Installation

When installed globally, you can use the setup script to create your .env.local file:

# If installed via pnpm
pnpm exec @bobmatnyc/ai-code-review/scripts/setup-env.js

# If installed via npm
npx @bobmatnyc/ai-code-review/scripts/setup-env.js

# If installed via Homebrew
node /opt/homebrew/lib/node_modules/@bobmatnyc/ai-code-review/scripts/setup-env.js

The script will guide you through setting up your API keys and will place the .env.local file in the correct directory.

Quick Fix for Environment Errors in Global Installation

If you're seeing errors about missing .env or .env.local files when running the globally installed tool, use this quick fix script:

# Download and run the fix script
curl -O https://raw.githubusercontent.com/bobmatnyc/ai-code-review/develop/scripts/global-env-fix.js
node global-env-fix.js

# You may need to run with sudo if permission errors occur
sudo node global-env-fix.js

This script will patch the global installation to avoid errors when environment files are missing and help you set up your API keys properly.

Alternatively, you can manually place your .env.local file in one of these locations:

If you installed via npm:

/path/to/global/node_modules/@bobmatnyc/ai-code-review/.env.local

If you installed via Homebrew:

/opt/homebrew/lib/node_modules/@bobmatnyc/ai-code-review/.env.local

Or, you can set the AI_CODE_REVIEW_DIR environment variable to specify the directory containing your .env.local:
```
export AI_CODE_REVIEW_DIR=/path/to/your/config/directory
```

Important: The tool first looks for .env.local in its own installation directory, not in the target project being reviewed. This allows you to keep your API keys in one place and review any project without modifying it.

Model configuration

Specify which model to use for code reviews using the format provider:model

The tool will automatically map this to the correct API model name

AI_CODE_REVIEW_MODEL=gemini:gemini-2.5-pro

or

AI_CODE_REVIEW_MODEL=openai:gpt-4o

or

AI_CODE_REVIEW_MODEL=anthropic:claude-4-opus

Optionally specify a different model for report consolidation/writing

This allows using a cheaper/faster model for the final report generation

AI_CODE_REVIEW_WRITER_MODEL=openai:gpt-4o-mini

See the Supported Models section for all available models and their API mappings

Custom context files

Comma-separated list of file paths to include as context for the code review

AI_CODE_REVIEW_CONTEXT=README.md,docs/architecture.md,src/types.ts

See the Supported Models section below for all available models


> Note: For backward compatibility, the tool also supports the old `CODE_REVIEW` prefix for environment variables, but the `AI_CODE_REVIEW` prefix is recommended.

## Supported Models

### Model Mapping System

The tool uses a centralized model mapping system that automatically converts user-friendly model names to provider-specific API formats. This ensures that you can use consistent model names across different providers without worrying about the specific API requirements of each provider.

For example, when you specify `anthropic:claude-3-opus` as your model, the tool automatically maps this to the correct API model name `claude-3-opus-20240229` when making requests to the Anthropic API.

The model mapping system provides the following benefits:

- **Consistent Model Names**: Use the same model naming convention across all providers
- **Automatic API Format Conversion**: No need to remember provider-specific model formats
- **Centralized Configuration**: All model mappings are defined in a single location for easy maintenance
- **Extensible**: New models can be added easily without changing the core code

You can see all available models and their mappings by running `ai-code-review --listmodels`.

### Google Models

| Model Name | Description | Context Window | API Key Required |
|------------|-------------|----------------|------------------|
| `gemini:gemini-2.5-pro` | Production-ready advanced reasoning model **(DEFAULT)** | 1M tokens | `AI_CODE_REVIEW_GOOGLE_API_KEY` |
| `gemini:gemini-2.5-pro-preview` | Most advanced reasoning and multimodal capabilities | 1M tokens | `AI_CODE_REVIEW_GOOGLE_API_KEY` |
| `gemini:gemini-2.0-flash` | Fast, efficient model with strong performance | 1M tokens | `AI_CODE_REVIEW_GOOGLE_API_KEY` |
| `gemini:gemini-2.0-flash-lite` | Ultra-fast, cost-efficient model | 1M tokens | `AI_CODE_REVIEW_GOOGLE_API_KEY` |
| `gemini:gemini-1.5-pro` | Previous generation large context model | 1M tokens | `AI_CODE_REVIEW_GOOGLE_API_KEY` |
| `gemini:gemini-1.5-flash` | Previous generation fast model | 1M tokens | `AI_CODE_REVIEW_GOOGLE_API_KEY` |

### OpenRouter Models

| Model Name | Description | Context Window | API Key Required |
|------------|-------------|----------------|------------------|
| `openrouter:anthropic/claude-4-opus` | Access Claude 4 Opus through OpenRouter | 200K tokens | `AI_CODE_REVIEW_OPENROUTER_API_KEY` |
| `openrouter:anthropic/claude-4-sonnet` | Access Claude 4 Sonnet through OpenRouter | 200K tokens | `AI_CODE_REVIEW_OPENROUTER_API_KEY` |
| `openrouter:openai/gpt-4o` | Access GPT-4o through OpenRouter | 128K tokens | `AI_CODE_REVIEW_OPENROUTER_API_KEY` |
| `openrouter:google/gemini-2.5-pro` | Access Gemini 2.5 Pro through OpenRouter | 1M tokens | `AI_CODE_REVIEW_OPENROUTER_API_KEY` |
| `openrouter:meta-llama/llama-3.3-70b` | Open source alternative | 131K tokens | `AI_CODE_REVIEW_OPENROUTER_API_KEY` |
| `openrouter:anthropic/claude-3-haiku` | Fast, affordable model | 200K tokens | `AI_CODE_REVIEW_OPENROUTER_API_KEY` |

### Anthropic Models

| Model Name | Description | Context Window | API Key Required |
|------------|-------------|----------------|------------------|
| `anthropic:claude-4-opus` | Most capable Claude model with superior reasoning | 200K tokens | `AI_CODE_REVIEW_ANTHROPIC_API_KEY` |
| `anthropic:claude-4-sonnet` | Balanced performance and cost for code review | 200K tokens | `AI_CODE_REVIEW_ANTHROPIC_API_KEY` |
| `anthropic:claude-3.5-sonnet` | Enhanced Claude 3 with improved capabilities | 200K tokens | `AI_CODE_REVIEW_ANTHROPIC_API_KEY` |
| `anthropic:claude-3-opus` | DEPRECATED: Migrate to Claude 4 Opus | 200K tokens | `AI_CODE_REVIEW_ANTHROPIC_API_KEY` |
| `anthropic:claude-3-sonnet` | Previous generation balanced model | 200K tokens | `AI_CODE_REVIEW_ANTHROPIC_API_KEY` |
| `anthropic:claude-3.5-haiku` | Fast, cost-effective model | 200K tokens | `AI_CODE_REVIEW_ANTHROPIC_API_KEY` |
| `anthropic:claude-3-haiku` | DEPRECATED: Migrate to Claude 3.5 Haiku | 200K tokens | `AI_CODE_REVIEW_ANTHROPIC_API_KEY` |

### OpenAI Models

| Model Name | Description | Context Window | API Key Required |
|------------|-------------|----------------|------------------|
| `openai:gpt-4o` | Multimodal model with vision capabilities | 128K tokens | `AI_CODE_REVIEW_OPENAI_API_KEY` |
| `openai:gpt-4.1` | Latest GPT-4 with improved reasoning | 128K tokens | `AI_CODE_REVIEW_OPENAI_API_KEY` |
| `openai:gpt-4.5` | DEPRECATED: Migrate to GPT-4.1 | 128K tokens | `AI_CODE_REVIEW_OPENAI_API_KEY` |
| `openai:gpt-4-turbo` | Fast GPT-4 variant | 128K tokens | `AI_CODE_REVIEW_OPENAI_API_KEY` |
| `openai:gpt-3.5-turbo` | Fast, cost-effective model | 16K tokens | `AI_CODE_REVIEW_OPENAI_API_KEY` |
| `openai:o3` | Advanced reasoning model | 100K tokens | `AI_CODE_REVIEW_OPENAI_API_KEY` |
| `openai:o3-mini` | Efficient reasoning model | 60K tokens | `AI_CODE_REVIEW_OPENAI_API_KEY` |

### Extending the Model Mapping System

If you need to add support for a new model or update an existing model mapping, you can modify the `MODEL_MAP` in `src/clients/utils/modelMaps.ts`. The model mapping system uses a simple key-value structure where the key is the user-friendly model name (e.g., `anthropic:claude-3-opus`) and the value contains the API-specific details:

```typescript
export const MODEL_MAP: Record<string, ModelMapping> = {
  'anthropic:claude-3-opus': {
    apiName: 'claude-3-opus-20240229',  // The actual API model name
    displayName: 'Claude 3 Opus',       // Human-readable name
    provider: 'anthropic',              // Provider identifier
    contextWindow: 200000,              // Context window size in tokens
    description: 'Claude 3 Opus - Anthropic\'s most powerful model',
    apiKeyEnvVar: 'AI_CODE_REVIEW_ANTHROPIC_API_KEY'  // Required API key
  },
  // Add more models here...
};

After adding a new model mapping, you can use it by setting the AI_CODE_REVIEW_MODEL environment variable to the new model key.

Model Mapping Utility Functions

The model mapping system provides several utility functions that you can use in your code:

// Get the API name for a model key
const apiName = getApiNameFromKey('anthropic:claude-3-opus');
// Returns: 'claude-3-opus-20240229'

// Get the full model mapping
const modelMapping = getModelMapping('anthropic:claude-3-opus');
// Returns: { apiName: 'claude-3-opus-20240229', displayName: 'Claude 3 Opus', ... }

// Get all models for a provider
const anthropicModels = getModelsByProvider('anthropic');
// Returns: ['anthropic:claude-3-opus', 'anthropic:claude-3-sonnet', ...]

// Parse a model string
const { provider, modelName } = parseModelString('anthropic:claude-3-opus');
// Returns: { provider: 'anthropic', modelName: 'claude-3-opus' }

// Get the full model key from provider and model name
const fullModelKey = getFullModelKey('anthropic', 'claude-3-opus');
// Returns: 'anthropic:claude-3-opus'

These utility functions make it easy to work with the model mapping system in your code.

AI-Generated Code Detection

The tool includes advanced AI detection capabilities specifically designed for coding test evaluations and hiring assessments. This feature analyzes code submissions to identify patterns that suggest AI assistance, helping maintain the integrity of technical assessments.

How AI Detection Works

The AI detection system uses a multi-analyzer approach that examines various aspects of code submissions:

Git History Analysis: Examines commit patterns, timing, and message structures that may indicate AI-generated content
Documentation Analysis: Analyzes README files, comments, and documentation for AI-typical patterns and structures
Structural Analysis: [Future] Evaluates code organization, naming conventions, and architectural patterns
Statistical Analysis: [Future] Performs linguistic analysis of variable names, comments, and coding style
Linguistic Analysis: [Future] Analyzes natural language patterns in comments and documentation

Using AI Detection

Basic AI Detection

# Enable AI detection for coding tests
ai-code-review ./candidate-submission --type coding-test --enable-ai-detection

# Use AI detection with custom confidence threshold
ai-code-review ./submission --type coding-test \
  --enable-ai-detection \
  --ai-detection-threshold 0.8

Advanced Configuration

# Specify which analyzers to use
ai-code-review ./submission --type coding-test \
  --enable-ai-detection \
  --ai-detection-analyzers git,documentation \
  --ai-detection-threshold 0.7

# Fail evaluation automatically if AI is detected
ai-code-review ./submission --type coding-test \
  --enable-ai-detection \
  --ai-detection-fail-on-detection

# Include detailed AI detection results in report
ai-code-review ./submission --type coding-test \
  --enable-ai-detection \
  --ai-detection-include-in-report

AI Detection Parameters

Parameter	Description	Default	Type
`--enable-ai-detection`	Enable AI-generated code detection	`false`	boolean
`--ai-detection-threshold`	Confidence threshold (0.0-1.0) for AI detection	`0.7`	number
`--ai-detection-analyzers`	Comma-separated list of analyzers to use	`git,documentation`	string
`--ai-detection-include-in-report`	Include AI detection results in the review report	`true`	boolean
`--ai-detection-fail-on-detection`	Automatically fail if AI-generated code is detected	`false`	boolean

Understanding AI Detection Results

Confidence Scores

0.9-1.0: Critical risk - Strong indicators of AI-generated content
0.8-0.9: High risk - Multiple AI patterns detected
0.6-0.8: Medium risk - Some concerning patterns
0.4-0.6: Low risk - Minor indicators present
0.0-0.4: Minimal risk - Likely human-authored

Pattern Types

High Confidence Patterns:

Large initial commits with complete functionality
Uniform commit timing patterns
Template-like README structure
Excessive or unusually uniform commenting

Medium Confidence Patterns:

Specific coding style patterns
Documentation structure anomalies
Variable naming conventions

Low Confidence Patterns:

Minor stylistic indicators
Edge case patterns

AI Detection in Different Scenarios

Hiring and Technical Assessments

# Strict evaluation for senior-level positions
ai-code-review ./candidate-submission --type coding-test \
  --enable-ai-detection \
  --ai-detection-threshold 0.6 \
  --ai-detection-fail-on-detection

# Detailed analysis for code review positions
ai-code-review ./submission --type coding-test \
  --enable-ai-detection \
  --ai-detection-analyzers git,documentation \
  --ai-detection-include-in-report

Educational Settings

# Moderate threshold for student assessments
ai-code-review ./student-project --type coding-test \
  --enable-ai-detection \
  --ai-detection-threshold 0.75 \
  --ai-detection-include-in-report

Code Review and Quality Assurance

# Light detection for internal code reviews
ai-code-review ./feature-branch --type coding-test \
  --enable-ai-detection \
  --ai-detection-threshold 0.8

Configuration File Support

You can configure AI detection in your .ai-code-review.yaml file:

# .ai-code-review.yaml
reviewType: coding-test
enableAiDetection: true
aiDetectionThreshold: 0.7
aiDetectionAnalyzers: "git,documentation"
aiDetectionIncludeInReport: true
aiDetectionFailOnDetection: false

# Assignment configuration
assignment:
  title: "Backend API Implementation"
  difficulty: "mid"
  type: "take-home"
  timeLimit: 240

# Evaluation criteria
criteria:
  correctness: 30
  codeQuality: 25
  architecture: 20
  performance: 15
  testing: 10

Best Practices for AI Detection

For Hiring Managers

Set Appropriate Thresholds: Use 0.7-0.8 for most positions, 0.6 for senior roles requiring high code authenticity
Combine with Interviews: Always follow up high-confidence detections with technical interviews
Consider Context: Factor in assignment complexity and time constraints
Document Decisions: Keep records of detection results and follow-up actions

For Educators

Educational Use: Use AI detection as a teaching tool to discuss code authenticity
Clear Policies: Establish clear guidelines about AI assistance in coursework
Progressive Thresholds: Use lower thresholds early in courses, higher for final projects
Student Discussion: Use detection results to facilitate conversations about learning

For Code Review

Team Standards: Establish team-wide policies about AI assistance in code development
Quality Assurance: Use detection to maintain code quality and authenticity standards
Process Integration: Integrate AI detection into existing code review workflows

Troubleshooting AI Detection

Common Issues

False Positives:

Very clean, well-structured code may trigger detection
Developers who follow strict coding standards consistently
Code based on well-known patterns or templates

False Negatives:

AI-generated code that has been significantly modified
Sophisticated AI tools that mimic human patterns
Mixed human-AI collaboration

Debugging Detection Results

# Enable debug logging for detailed analysis
ai-code-review ./submission --type coding-test \
  --enable-ai-detection \
  --debug

# Use multiple analyzers for comprehensive analysis
ai-code-review ./submission --type coding-test \
  --enable-ai-detection \
  --ai-detection-analyzers git,documentation,structural

Integration Examples

CI/CD Pipeline Integration

# GitHub Actions example
name: Code Review with AI Detection
on: [pull_request]

jobs:
  ai-review:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
        with:
          fetch-depth: 0  # Important for git history analysis
      
      - name: AI Code Review with Detection
        run: |
          npx @bobmatnyc/ai-code-review ./src \
            --type coding-test \
            --enable-ai-detection \
            --ai-detection-threshold 0.8 \
            --format json \
            --output ai-review-results.json
      
      - name: Check AI Detection Results
        run: |
          if grep -q '"isAIGenerated": true' ai-review-results.json; then
            echo ":⚠️:AI-generated code detected"
            exit 1
          fi

Batch Processing

#!/bin/bash
# Process multiple candidate submissions

for submission in ./candidates/*/; do
  candidate=$(basename "$submission")
  echo "Analyzing $candidate..."
  
  ai-code-review "$submission" \
    --type coding-test \
    --enable-ai-detection \
    --ai-detection-threshold 0.7 \
    --format json \
    --output "results/${candidate}-analysis.json"
done

# Generate summary report
echo "Generating summary report..."
node scripts/generate-summary-report.js results/

Security and Privacy Considerations

Data Privacy: AI detection analysis is performed locally; no code is sent to external services
False Positives: Always verify high-confidence detections through additional assessment methods
Legal Compliance: Ensure AI detection usage complies with local employment and education laws
Transparency: Consider disclosing AI detection usage to candidates and students

Supported File Types

AI detection works best with:

Git repositories with commit history
Projects with README files and documentation
Standard project structures (package.json, etc.)
Multiple file types for comprehensive analysis

Code Tracing for Unused Code Detection

The tool includes a powerful code tracing feature for identifying unused code with high confidence. This feature uses a multi-pass approach to analyze code dependencies and references, providing detailed evidence for each element identified as unused.

How Code Tracing Works

The code tracing feature follows a comprehensive approach:

Entry Point & Dependency Mapping: Identifies all entry points to the codebase and maps module dependencies
Reference Tracing: Finds all references to each code element throughout the codebase
Verification & Confidence Assessment: Evaluates evidence and assigns confidence levels (high, medium, low)

Using Code Tracing

To enable code tracing:

# Basic unused code detection
ai-code-review src --type unused-code

# Enhanced detection with deep code tracing
ai-code-review src --type unused-code --trace-code

Code Tracing Benefits

High Confidence Detection: Thorough evidence collection ensures recommendations are reliable
Detailed Evidence: Each element includes complete evidence chain showing why it's unused
Risk Assessment: Evaluates potential risks of removing each element
Removal Scripts: Automatically generates scripts for safely removing unused code
Edge Case Detection: Considers special cases like dynamic imports and reflection patterns

Confidence Levels

Code tracing assigns confidence levels to each finding:

High: Clear evidence the element is never referenced (safe to remove)
Medium: Likely unused but with some uncertainty (verify before removing)
Low: Possibly unused but with significant uncertainty (requires further investigation)

Testing API Connections

You can test your API connections to verify that your API keys are valid and working correctly:

# Test the configured model
ai-code-review test-model

# Test a specific model
ai-code-review test-model --model anthropic:claude-4-sonnet

# Test API connections before running a review
ai-code-review src --test-api

This will test connections to the configured API providers and provide detailed feedback on the status of each connection.

Requirements

Node.js >= 16.0.0
Google Generative AI API key or OpenRouter API key

License

MIT