DiffScope

A composable code review engine for automated diff analysis.

Features

• Model Agnostic: Works with OpenAI, Anthropic Claude 4, Ollama, and any OpenAI-compatible API
• Git Integration: Review uncommitted, staged, or branch changes directly
• PR Reviews: Analyze and comment on GitHub pull requests
• Smart Prompting: Advanced prompt engineering with examples, XML structure, and chain-of-thought
• Commit Messages: AI-powered commit message suggestions following conventional commits
• Composable Architecture: Modular components that work together
• Plugin System: Extensible pre-analyzers and post-processors
• Multiple Outputs: JSON, patch, or markdown formats
• CI/CD Ready: GitHub Action, GitLab CI, and Docker support
• Smart Review: Enhanced analysis with confidence scoring, fix effort estimation, and executive summaries

Quick Start

Install via Homebrew (macOS/Linux)

brew tap haasonsaas/diffscope
brew install diffscope

Install from source

cargo install --git https://github.com/Haasonsaas/diffscope

Docker

docker run --rm -v $(pwd):/workspace ghcr.io/haasonsaas/diffscope:latest review --diff /workspace/pr.diff

Usage

Review a diff file

diffscope review --diff changes.diff

Review from stdin

git diff | diffscope review

Git integration

# Review uncommitted changes
diffscope git uncommitted

# Review staged changes
diffscope git staged

# Review changes from a branch
diffscope git branch main

# Suggest commit message for staged changes
diffscope git suggest

Pull Request review

# Review current PR
diffscope pr

# Review specific PR
diffscope pr --number 123

# Post comments directly to PR
diffscope pr --post-comments

Smart Review (Enhanced Analysis)

# Comprehensive analysis with executive summary
diffscope smart-review --diff changes.diff

# Enhanced review from stdin  
git diff | diffscope smart-review

# Save detailed report to file
diffscope smart-review --diff pr.patch --output detailed-report.md

Use different models

# OpenAI GPT-4
export OPENAI_API_KEY=your-openai-key
diffscope review --model gpt-4o --diff changes.diff

# Anthropic Claude
export ANTHROPIC_API_KEY=your-anthropic-key
diffscope review --model claude-3-5-sonnet-20241022 --diff changes.diff
diffscope review --model claude-3-haiku-20240307 --diff changes.diff

# Local Ollama
diffscope review --model ollama:codellama --diff changes.diff
diffscope review --model ollama:llama3.2 --diff changes.diff

Supported Models

OpenAI: gpt-4o, gpt-4-turbo, gpt-3.5-turbo

Anthropic:

• Claude 4: claude-opus-4-20250514, claude-sonnet-4-20250514
• Claude 3.5: claude-3-5-sonnet-20241022, claude-3-5-haiku-20240307
• Claude 3: claude-3-opus-20240229, claude-3-haiku-20240307

Ollama: Any locally installed model (codellama, llama3.2, mistral, etc.)

Output formats

# JSON (default)
diffscope review --diff changes.diff --output-format json

# Markdown
diffscope review --diff changes.diff --output-format markdown > review.md

# Patch comments
diffscope review --diff changes.diff --output-format patch

GitHub Action

name: Code Review
on: [pull_request]

jobs:
  review:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: haasonsaas/diffscope@v1
        with:
          model: gpt-4o
          openai-api-key: ${{ secrets.OPENAI_API_KEY }}

Configuration

Create a .diffscope.yml file in your repository:

model: gpt-4o
temperature: 0.2
max_tokens: 4000
system_prompt: |
  You are an expert code reviewer. Focus on:
  - Security vulnerabilities
  - Performance issues
  - Best practices
  - Code clarity

Plugin Development

Create custom analyzers:

export interface PreAnalyzer {
  id: string
  run(diff: UnifiedDiff, repoPath: string): Promise<LLMContextChunk[]>
}

export interface PostProcessor {
  id: string
  run(comments: Comment[], repoPath: string): Promise<Comment[]>
}

Architecture

graph LR
  A[git diff] --> B(core-engine)
  subgraph core-engine
    B1[Diff Parser]
    B2[Context Fetcher]
    B3[Prompt Builder]
    B4[LLM Adapter]
    B5[Comment Synthesizer]
  end
  B -->|JSON| C(output)

License

Apache-2.0 License. See LICENSE for details.

Example Output

Standard Review

Line 32: Security - Logging raw input data may expose sensitive information. 
Risk of data leakage. Remove or sanitize before logging.

Line 14: Bug - TODO indicates missing error handling. Could cause crashes. 
Implement proper error handling.

Smart Review Output

# 🤖 Smart Review Analysis Results

## 📊 Executive Summary

🟡 **Code Quality Score:** 7.2/10
📝 **Total Issues Found:** 3
🚨 **Critical Issues:** 1
📁 **Files Analyzed:** 2

### 📈 Issue Breakdown
| Severity | Count | Category | Count |
|----------|-------|----------|-------|
| Error | 1 | Security | 1 |
| Warning | 2 | Performance | 1 |

### 🎯 Priority Actions
1. Address 1 security issue(s) immediately
2. Consider a performance audit - multiple optimization opportunities found

### 🔴 Critical Issues (Fix Immediately)

#### 🔒 **auth.py:42** - 🔴 Significant Effort Security
**Confidence:** 95% | **Tags:** `security`, `sql`, `injection`

SQL injection vulnerability in user authentication. The query directly interpolates user input without parameterization.

**💡 Recommended Fix:**
Use parameterized queries: `query = "SELECT * FROM users WHERE username=%s AND password=%s"`

**🔧 Code Example:**
```diff
- query = f"SELECT * FROM users WHERE username='{username}' AND password='{password}'"
+ query = "SELECT * FROM users WHERE username=%s AND password=%s"
+ cursor.execute(query, (username, password))

### Commit Message Suggestion

feat(auth): add JWT-based authentication system

## Author

Jonathan Haas <jonathan@haas.holdings>

## Contributing

Contributions are welcome! Please open an issue first to discuss what you would like to change.

## Support

- GitHub Issues: [github.com/Haasonsaas/diffscope/issues](https://github.com/Haasonsaas/diffscope/issues)