|
| 1 | +# Diff Hound |
| 2 | + |
| 3 | +Diff Hound is an automated AI-powered code review tool that posts intelligent, contextual comments directly on pull requests across supported platforms. |
| 4 | + |
| 5 | +Supports GitHub today. GitLab and Bitbucket support are planned. |
| 6 | + |
| 7 | +--- |
| 8 | + |
| 9 | +## ✨ Features |
| 10 | + |
| 11 | +- 🧠 Automated code review using OpenAI (Upcoming: Claude, DeepSeek, CodeLlama) |
| 12 | +- 💬 Posts inline or summary comments on pull requests |
| 13 | +- 🔌 Plug-and-play architecture for models and platforms |
| 14 | +- ⚙️ Configurable with JSON/YAML config files and CLI overrides |
| 15 | +- 🛠️ Designed for CI/CD pipelines and local runs |
| 16 | +- 🧐 Tracks last reviewed commit to avoid duplicate reviews |
| 17 | + |
| 18 | +--- |
| 19 | + |
| 20 | +## 🛠️ Installation |
| 21 | + |
| 22 | +### Option 1: Install via npm |
| 23 | + |
| 24 | +```bash |
| 25 | +npm install -g diff-hound |
| 26 | +``` |
| 27 | + |
| 28 | +### Option 2: Install from source |
| 29 | + |
| 30 | +```bash |
| 31 | +git clone https://github.com/runtimebug/diff-hound.git |
| 32 | +cd diff-hound |
| 33 | +npm install |
| 34 | +npm run build |
| 35 | +npm link |
| 36 | +``` |
| 37 | + |
| 38 | +--- |
| 39 | + |
| 40 | +## 🚀 How to Use |
| 41 | + |
| 42 | +### Step 1: Setup Environment Variables |
| 43 | + |
| 44 | +Copy the provided `.env.example` to `.env` and fill in your credentials: |
| 45 | + |
| 46 | +```bash |
| 47 | +cp .env.example .env |
| 48 | +``` |
| 49 | + |
| 50 | +Then modify with your keys / tokens: |
| 51 | + |
| 52 | +```env |
| 53 | +# Platform tokens |
| 54 | +GITHUB_TOKEN=your_github_token # Requires 'repo' scope |
| 55 | +
|
| 56 | +# AI Model API keys |
| 57 | +OPENAI_API_KEY=your_openai_key |
| 58 | +``` |
| 59 | + |
| 60 | +> 🔐 `GITHUB_TOKEN` is used to fetch PRs and post comments – [get it here](https://github.com/settings/personal-access-tokens) |
| 61 | +> 🔐 `OPENAI_API_KEY` is used to generate code reviews via GPT – [get it here](https://platform.openai.com/api-keys) |
| 62 | +
|
| 63 | +--- |
| 64 | + |
| 65 | +### Step 2: Create a Config File |
| 66 | + |
| 67 | +You can define your config in `.aicodeconfig.json` or `.aicode.yml`: |
| 68 | + |
| 69 | +#### JSON Example (`.aicodeconfig.json`) |
| 70 | + |
| 71 | +```json |
| 72 | +{ |
| 73 | + "provider": "openai", |
| 74 | + "model": "gpt-4o", // Or any other openai model |
| 75 | + "endpoint": "", // Optional: custom endpoint |
| 76 | + "gitProvider": "github", |
| 77 | + "repo": "your-username/your-repo", |
| 78 | + "dryRun": false, |
| 79 | + "verbose": false, |
| 80 | + "rules": [ |
| 81 | + "Prefer const over let when variables are not reassigned", |
| 82 | + "Avoid reassigning const variables", |
| 83 | + "Add descriptive comments for complex logic", |
| 84 | + "Remove unnecessary comments", |
| 85 | + "Follow the DRY (Don't Repeat Yourself) principle", |
| 86 | + "Use descriptive variable and function names", |
| 87 | + "Handle errors appropriately", |
| 88 | + "Add type annotations where necessary" |
| 89 | + ], |
| 90 | + "ignoreFiles": ["*.md", "package-lock.json", "yarn.lock", "LICENSE", "*.log"], |
| 91 | + "commentStyle": "inline", |
| 92 | + "severity": "suggestion" |
| 93 | +} |
| 94 | +``` |
| 95 | + |
| 96 | +#### YAML Example (`.aicode.yml`) |
| 97 | + |
| 98 | +```yaml |
| 99 | +provider: openai |
| 100 | +model: gpt-4o # Or any other openai model |
| 101 | +endpoint: "" # Optional: custom endpoint |
| 102 | +gitProvider: github |
| 103 | +repo: your-username/your-repo |
| 104 | +dryRun: false |
| 105 | +verbose: false |
| 106 | +commentStyle: inline |
| 107 | +severity: suggestion |
| 108 | +ignoreFiles: |
| 109 | + - "*.md" |
| 110 | + - package-lock.json |
| 111 | + - yarn.lock |
| 112 | + - LICENSE |
| 113 | + - "*.log" |
| 114 | +rules: |
| 115 | + - Prefer const over let when variables are not reassigned |
| 116 | + - Avoid reassigning const variables |
| 117 | + - Add descriptive comments for complex logic |
| 118 | + - Remove unnecessary comments |
| 119 | + - Follow the DRY (Don't Repeat Yourself) principle |
| 120 | + - Use descriptive variable and function names |
| 121 | + - Handle errors appropriately |
| 122 | + - Add type annotations where necessary |
| 123 | +``` |
| 124 | +
|
| 125 | +--- |
| 126 | +
|
| 127 | +### Step 3: Run It |
| 128 | +
|
| 129 | +```bash |
| 130 | +diff-hound |
| 131 | +``` |
| 132 | + |
| 133 | +Or override config values via CLI: |
| 134 | + |
| 135 | +```bash |
| 136 | +diff-hound --repo=owner/repo --provider=openai --model=gpt-4o --dry-run |
| 137 | +``` |
| 138 | + |
| 139 | +> Add `--dry-run` to **print comments to console** instead of posting them. |
| 140 | +
|
| 141 | +--- |
| 142 | + |
| 143 | +### Output Example (Dry Run) |
| 144 | + |
| 145 | +```bash |
| 146 | +== Comments for PR #42: Fix input validation == |
| 147 | + |
| 148 | +src/index.ts:17 — |
| 149 | +Prefer `const` over `let` since `userId` is not reassigned. |
| 150 | + |
| 151 | +src/utils/parse.ts:45 — |
| 152 | +Consider refactoring to reduce nesting. |
| 153 | +``` |
| 154 | + |
| 155 | +--- |
| 156 | + |
| 157 | +### Optional CLI Flags |
| 158 | + |
| 159 | +| Flag | Short | Description | |
| 160 | +| ------------------ | ----- | --------------------------------------- | |
| 161 | +| `--provider` | `-p` | AI model provider (e.g. `openai`) | |
| 162 | +| `--model` | `-m` | AI model (e.g. `gpt-4o`, `gpt-4`, etc.) | |
| 163 | +| `--model-endpoint` | `-e` | Custom API endpoint for the model | |
| 164 | +| `--git-provider` | `-g` | Repo platform (default: `github`) | |
| 165 | +| `--repo` | `-r` | GitHub repo in format `owner/repo` | |
| 166 | +| `--comment-style` | `-s` | `inline` or `summary` | |
| 167 | +| `--dry-run` | `-d` | Don’t post comments, only print | |
| 168 | +| `--verbose` | `-v` | Enable debug logs | |
| 169 | +| `--config-path` | `-c` | Custom config file path | |
| 170 | + |
| 171 | +--- |
| 172 | + |
| 173 | +## 🛠️ Development |
| 174 | + |
| 175 | +### Project Structure |
| 176 | + |
| 177 | +``` |
| 178 | +diff-hound/ |
| 179 | +├── bin/ # CLI entrypoint |
| 180 | +├── src/ |
| 181 | +│ ├── cli/ # CLI argument parsing |
| 182 | +│ ├── config/ # JSON/YAML config handling |
| 183 | +│ ├── core/ # Diff parsing, formatting |
| 184 | +│ ├── models/ # AI model adapters |
| 185 | +│ ├── platforms/ # GitHub, GitLab, etc. |
| 186 | +│ └── types/ # TypeScript types |
| 187 | +├── .env |
| 188 | +├── README.md |
| 189 | +``` |
| 190 | + |
| 191 | +--- |
| 192 | + |
| 193 | +### Add Support for New AI Models |
| 194 | + |
| 195 | +Create a new class in `src/models/` that implements the `CodeReviewModel` interface. |
| 196 | + |
| 197 | +--- |
| 198 | + |
| 199 | +### Add Support for New Platforms |
| 200 | + |
| 201 | +Create a new class in `src/platforms/` that implements the `CodeReviewPlatform` interface. |
| 202 | + |
| 203 | +--- |
| 204 | + |
| 205 | +## ✅ Next Steps |
| 206 | + |
| 207 | +🔧 Add Winston for production-grade logging |
| 208 | +🌐 Implement GitLab and Bitbucket platform adapters |
| 209 | +🌍 Add support for other AI model providers (e.g. Anthropic, DeepSeek...) |
| 210 | +💻 Add support for running local models (e.g. Ollama, Llama.cpp, Hugging Face transformers) |
| 211 | +📤 Add support for webhook triggers (e.g., GitHub Actions, GitLab CI) |
| 212 | +🧪 Add unit and integration test suites (Jest or Vitest) |
| 213 | +📦 Publish Docker image for CI/CD use |
| 214 | +🧩 Enable plugin hooks for custom rule logic |
| 215 | +🗂 Add support for reviewing diffs from local branches or patch files |
| 216 | + |
| 217 | +--- |
| 218 | + |
| 219 | +## 📜 License |
| 220 | + |
| 221 | +MIT – Use freely, contribute openly. |
0 commit comments