Documentry

Documentry is a AI-powered Typescript library that uses LLM models to understand your Next.js API routes and automatically generate detailed OpenAPI documentation in multiple formats: json, yaml, and interactive html.

With a single terminal command, Documentry scans every API route in your Next.js project, understand the actual code of your route.ts files, and generates a valid OpenAPI Specification (OAS) file that describes your endpoints.

Key Features

• 🚀 Automatically scams your project and detects your Next.js API routes
• 🧠 Uses AI to understand the actual code of your routes
• 📝 Creates OpenAPI 3.0 specifications in json, yaml, or interactive html format
• 🔄 Currently supports OpenAI and Anthropic models

Installation

npm install documentry --save-dev

Usage

Command line

npx documentry 

Programmatic API

import { Documentry } from 'documentry'

// Create a new Documentry instance
const documentry = new Documentry()

// Generate OpenAPI specs
await documentry.generate()

Full usage example

const documentry = new Documentry({
  provider: 'anthropic',
  model: 'claude-3-5-sonnet-latest',
  apiKey: process.env.ANTHROPIC_API_KEY,
  dir: './app/api',
  routes: ['/user', '/products/*'],
  outputFile: './docs/openapi',
  format: 'html', // 'yaml', 'json', or 'html'
  info: {
    title: 'My API',
    version: '1.0.0',
    description: 'My API description'
  },
  servers: [
    {
      url: 'http://localhost:3000/api',
      description: 'Local server'
    },
    {
      url: 'https://api.example.com',
      description: 'Production server'
    }
  ]
})

await documentry.generate()

An example of the generated OpenAPI documentation in HTML format:

More examples can be found in the examples directory.

Environment Variables

You can configure the LLM settings with an .env file:

LLM_PROVIDER=your-llm-provider # openai or anthropic; defaults to anthropic
LLM_MODEL=your-llm-model # defaults to claude-3-5-sonnet-latest
ANTHROPIC_API_KEY=your-anthropic-key
OPENAI_API_KEY=your-openai-key

Configuration Options

The CLI usage supports the following options:

Flag	Description	Default
--dir <directory>	Root directory for your Nextjs API routes (./app/api, ./src/app/api, etc)	./app/api
--routes <routes>	List of routes to process (e.g., "/user,/products/*")	All routes are considered
--servers <servers>	List of server URLs (e.g. "url1|description1, url2...")	URL: http://localhost:3000/api
-o, --output-file <file>	Output folder/file for the generated OpenAPI specs	./docs/openapi
-f, --format	The format for the generated OpenAPI file (yaml, json, or html)	yaml
-t, --title <title>	Title for the OpenAPI spec	Next.js API
-d, --description <description>	Description for the OpenAPI spec	API documentation for Next.js routes
-v, --version <version>	Version for the OpenAPI spec	1.0.0
-p, --provider <provider>	LLM provider (anthropic or openai)	Env variable LLM_PROVIDER
-m, --model <model>	LLM model to use	Env variable LLM_MODEL
-k, --api-key <key>	LLM provider API key	Env variable ANTHROPIC_API_KEY or OPENAI_API_KEY, according to the provider

Development

Prerequisites

• Node.js >= 14.0.0
• npm >= 6.0.0

Setting up the Development Environment

1. Clone the repository:

   git clone https://github.com/thiagobarbosa/documentry
   cd documentry

2. Install dependencies:

   npm install

3. Build the project:

   npm run build

4. Run in development mode:

   npm run dev

License

This project is licensed under the MIT License.