manque-ai

✨ Features

• 🚀 Dual Mode Operation: Run as a GitHub Action or a local CLI tool.
• 🤖 Multi-Provider LLM Support: First-class support for OpenAI, Anthropic Claude, Google Gemini, and OpenRouter.
• 🧠 Intelligent Analysis: Generates executive summaries, walkthroughs, and line-by-line review comments.
• 🔒 Security First: dedicated analysis for hardcoded secrets and potential vulnerabilities.
• 💻 Local Pre-PR Checks: Review your code locally before you even push.
• 🎨 Custom Styling: Enforce your team's unique style guide and best practices.

💻 Local Development (Pre-PR Check)

Review your changes locally without pushing to GitHub. This is perfect for catching issues early!

1. Installation

Quick Install (Recommended)

./install.sh

Manual Install

go install github.com/igcodinap/manque-ai@latest
# or build from source
git clone https://github.com/igcodinap/manque-ai
cd manque-ai && go build -o manque-ai .

2. Updating

Easily update to the latest version:

manque-ai update

2. Setup (One-time)

You can set your LLM credentials as environment variables or using a .env file in the project root. Note: GH_TOKEN is OPTIONAL for local runs!

Option A: Using a .env file (Recommended)

Copy the example file and fill in your keys:

cp .env.example .env
# Edit .env and add your LLM_API_KEY

Option B: Exporting variables (or adding to .env)

OpenAI

export LLM_PROVIDER=openai
export LLM_API_KEY=sk-...

Anthropic

export LLM_PROVIDER=anthropic
export LLM_API_KEY=sk-ant-...

Local Ollama (No key required, just pointing to your local instance)

export LLM_PROVIDER=openai
export LLM_BASE_URL=http://localhost:11434/v1
export LLM_API_KEY=ollama
export LLM_MODEL=llama3 # Make sure to `ollama pull llama3` first!

OpenRouter

export LLM_PROVIDER=openrouter
export LLM_API_KEY=sk-or-...
export LLM_MODEL=anthropic/claude-3.5-sonnet

3. Run Review

# Review changes in your current branch vs main
manque-ai local

# Compare specific branches
manque-ai local --base develop --head feature-login

# Debug mode (see exact API calls and diff sizes)
manque-ai local --debug

🚀 GitHub Action Usage

Integrate directly into your CI/CD pipeline to review every Pull Request automatically.

name: AI Code Review
on:
  pull_request:
    types: [opened, synchronize]

jobs:
  review:
    runs-on: ubuntu-latest
    permissions:
      contents: read
      pull-requests: write
    steps:
      - name: Manque AI
        uses: docker://ghcr.io/igcodinap/manque-ai:latest
        env:
          GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          LLM_API_KEY: ${{ secrets.OPENAI_API_KEY }}
          LLM_PROVIDER: "openai"
          LLM_MODEL: "gpt-4o"

Configuration Options

🛠️ Advanced CLI Usage

The CLI can also be used to review remote PRs or check GitHub Actions context.

# Review a specific remote PR
manque-ai --repo owner/repo --pr 123

# Review by URL
manque-ai --url https://github.com/owner/repo/pull/123

🧠 Architecture

The project is built with modularity in mind, separating the "brain" from the interface.

├── cmd/               # CLI Commands
│   ├── root.go        # GitHub Action / Remote Review
│   └── local.go       # Local Pre-PR Review
├── pkg/
│   ├── review/        # Core Review Engine (Shared Logic)
│   ├── ai/            # LLM Client Adapters
│   ├── diff/          # Git Diff Parser
│   └── github/        # GitHub API Client
└── internal/          # Config & Logging

🤝 Contributing

We love contributions! Please fork the repository and submit a Pull Request.

📄 License

MIT Licensed. default_api.