llm-blog-seo

A powerful Node.js package that leverages LLMs (Large Language Models) and SEO best practices to automatically generate high-quality blog posts with proper research and citations.

Features

• 🤖 Support for multiple LLM providers (OpenAI, DeepSeek, with more coming soon)
• 🔍 Integrated research capabilities using Tavily API
• ✍️ Automatic blog post generation with customizable parameters
• 📊 SEO-optimized content with keyword targeting
• 🎯 Customizable writing styles and target audiences
• 📝 Automatic outline generation and content structuring
• 📚 Proper citations and references

Installation

npm install llm-blog-seo

Quick Start

import { initialize, BlogGenerationConfig } from "llm-blog-seo";
// Initialize the package with your configuration
const generator = initialize({
  llm: {
    provider: "openai",
    apiKey: "your-openai-api-key",
  },
  tavily: {
    apiKey: "your-tavily-api-key",
  },
});
// Generate a blog post
const config: BlogGenerationConfig = {
  topic: "The Future of Artificial Intelligence",
  targetWordCount: 1500,
  style: "technical",
  targetAudience: "tech professionals",
  keywords: ["AI", "machine learning", "future technology"],
};
const blogPost = await generator.generateBlog(config);
console.log(blogPost);

Configuration

LLM Configuration

The package supports various LLM providers. Here's how to configure them:

{
llm: {
provider: 'openai' | 'deepseek' | 'anthropic' | 'custom',
apiKey: string,
model?: string,
temperature?: number,
maxTokens?: number,
baseUrl?: string,
customHeaders?: Record<string, string>
}
}

Tavily Search Configuration

{
tavily: {
apiKey: string,
maxResults?: number,
searchDepth?: 'basic' | 'advanced'
}
}

Default Settings

{
defaults: {
style?: 'formal' | 'casual' | 'technical',
targetAudience?: string,
targetWordCount?: number
}
}

Blog Generation Options

| Option | Type | Description | | --------------- | ----------------------------------- | --------------------------------- | | topic | string | The main topic of the blog post | | targetWordCount | number | Desired length of the blog post | | style | 'formal' | 'casual' | 'technical' | Writing style | | targetAudience | string | Intended audience for the content | | keywords | string[] | Target keywords for SEO |

Response Format

The generated blog post will have the following structure:

interface BlogPost {
  title: string;
  content: string;
  metadata: {
    wordCount: number;
    keywords: string[];
    references: string[];
  };
}

Examples

See the example.ts file for more detailed usage examples.

Requirements

• Node.js 16.x or higher
• Valid API keys for your chosen LLM provider
• Tavily API key for research capabilities

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Support

If you encounter any issues or have questions, please file an issue on the GitHub repository.

Acknowledgments

• OpenAI for their GPT models
• Tavily for their research API
• All contributors who help improve this package

Architecture

graph TB
    subgraph Client["Client Application"]
        A[Initialize Config] --> B[BlogGenerator]
    end

    subgraph ConfigSystem["Configuration System"]
        C[ConfigManager<br>Singleton] --> D[LLMBlogSEOConfig]
    end

    subgraph BlogGeneration["Blog Generation System"]
        B --> E[SearchService<br>Search Agent]
        B --> F[ContentGenerator<br>Content Agent]

        E --> |Search Results| F

        subgraph ContentGenFlow["Content Generation Flow"]
            F --> G[Context Preparation<br>prepareContext]
            G --> H[Outline Generator<br>generateOutline]
            H --> I[Post Generator<br>generateFullPost]
        end
    end

    subgraph LLMSystem["LLM System"]
        J[LLMProviderFactory<br>Factory Pattern] --> K[LLM Providers]

        subgraph Providers["LLM Providers"]
            K --> L[OpenAI Provider]
            K --> M[Deepseek Provider]
            K --> N[Other Providers...]
        end
    end

    subgraph SearchSystem["Search System"]
        E --> O[Tavily API<br>Search Service]
        O --> P[Content Extraction Service]
    end

    F --> |Returns| SearchResults
    F --> |Generates| BlogPost

    %% Modern color scheme
    classDef system fill:#e3f2fd,stroke:#2196f3,stroke-width:2px
    classDef agent fill:#f3e5f5,stroke:#9c27b0,stroke-width:2px
    classDef service fill:#e8f5e9,stroke:#4caf50,stroke-width:2px

    class ConfigSystem,LLMSystem,SearchSystem,BlogGeneration system
    class E,F,G,H,I agent
    class O,P,L,M,N service