janusdoc

Janusdoc

Installation • Quick Start • Commands • CI/CD • Evaluation Suite

AI-powered documentation update suggester for pull requests. JanusDoc analyzes code changes in your PRs and automatically suggests documentation updates based on your project's existing docs and style guide.

Installation

npm install -g janusdoc

Or use with npx:

npx janusdoc [command]

Quick Start

• Initialize JanusDoc in your project:

  janusdoc init
  

• Analyze a pull request:

  janusdoc run --pr 123 --repo owner/repo
  

Commands

janusdoc init

Initialize JanusDoc in your current project. This command will:

• Detect or create your documentation directory
• Generate a .janusdoc.json configuration file
• Scan your existing documentation
• Generate an AI-powered style guide based on your docs
• Create embeddings for semantic search

Options:

• -d, --docs-path <path> - Path to documentation directory (default: auto-detected or "docs")

Example:

janusdoc init
janusdoc init --docs-path documentation

Environment Variables:

• OPENAI_API_KEY - Required for AI-powered features (style guide generation and semantic search)

janusdoc run

Analyze a pull request and suggest documentation updates. Posts a comment on the PR with suggestions if documentation updates are needed.

Required Options:

• -p, --pr <number> - Pull request number
• -r, --repo <owner/repo> - Repository in owner/repo format

Optional:

• -t, --token <token> - GitHub token (defaults to GITHUB_TOKEN environment variable)

Example:

# Using GITHUB_TOKEN from environment
janusdoc run --pr 42 --repo myorg/myproject

# Providing token explicitly
janusdoc run --pr 42 --repo myorg/myproject --token ghp_xxxxx

Environment Variables:

• GITHUB_TOKEN - GitHub personal access token with repo access
• OPENAI_API_KEY - Required for AI-powered analysis

Configuration

After running janusdoc init, a .janusdoc.json file is created:

{
  "docsPath": "docs"
}

JanusDoc also creates a .janusdoc/ directory containing:

• auto_styleguide.md - Auto-generated documentation style guide (can be customized)
• embeddings.json - Vector embeddings for semantic search

How It Works

• Initialization: JanusDoc scans your documentation, learns your style, and creates embeddings for efficient searching
• PR Analysis: When analyzing a PR, JanusDoc:
  • Fetches the code changes from GitHub
  • Summarizes the changes using AI
  • Uses semantic search to find relevant documentation
  • Analyzes whether documentation updates are needed
  • Posts suggestions as a PR comment

CI/CD Integration

Add JanusDoc to your GitHub Actions workflow:

name: Documentation Check
on: [pull_request]

jobs:
  docs-check:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0

      - uses: actions/setup-node@v4
        with:
          node-version: "20"

      - name: Run JanusDoc
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
        run: |
          npx janusdoc run --pr ${{ github.event.pull_request.number }} --repo ${{ github.repository }}

Requirements

• Node.js 18 or higher
• GitHub personal access token with repo scope
• OpenAI API key for AI-powered features

License

https://osaasy.dev/