Article Rewriter LangGraph Agent

Standalone LangGraph agent for article rewriting with browser capabilities. This production-ready agent extracts content from URLs, generates SEO-optimized articles, and humanizes the content using advanced writing style guidelines.

Features

• Web Content Extraction: Automatically scrapes and extracts content from URLs
• SEO-Optimized Article Generation: Creates articles with target keywords, readability levels, and SEO scores
• Content Humanization: Applies advanced writing style guidelines for natural, engaging content
• Structured Output: Saves article content and metadata in separate, organized files

Workflow

The agent follows a streamlined workflow:

1. Extract Content (if URL provided) - Scrapes and extracts content from the given URL
2. Generate Article - Creates an SEO-optimized article based on keywords, tone, and target audience
3. Humanize Content - Applies writing style guidelines to make content more natural and engaging

Installation

Install dependencies from the requirements file:

pip install -r requirements.txt

Required Environment Variables

Required: OpenAI API Key

Set your OpenAI API key using one of these methods:

Option 1: Environment Variable

export OPENAI_API_KEY=your_openai_api_key_here

Option 2: .env File

Create a .env file in the same directory as the script:

OPENAI_API_KEY=your_openai_api_key_here

The script will automatically load the .env file if python-dotenv is installed.

Optional: LangSmith for LLM Logging and Observability

To view detailed LLM logs, traces, and monitor agent performance, add LangSmith environment variables to your .env file:

# LangSmith Configuration (Optional)
LANGCHAIN_TRACING_V2=true
LANGCHAIN_ENDPOINT=https://api.smith.langchain.com
LANGCHAIN_API_KEY=your_langsmith_api_key_here
LANGCHAIN_PROJECT=article-rewriter-agent

Getting Your LangSmith API Key:

1. Sign up for a free account at https://smith.langchain.com
2. Navigate to Settings → API Keys
3. Create a new API key
4. Copy the key to your .env file

Benefits of LangSmith:

• View all LLM requests and responses in real-time
• Monitor token usage and costs
• Debug agent workflow execution
• Track performance metrics
• Analyze prompt effectiveness

If LangSmith is not configured, the agent will work normally but without observability features.

Usage

Run as Standalone Script

python article_rewriter_agent.py

Or run as a module:

python -m article_rewriter_agent

Customize the Example

Edit the example_usage() function in the script to customize:

• URL: The source URL to extract content from
• Primary Keywords: Main SEO keywords for the article
• Secondary Keywords: Additional keywords (optional)
• Tone: Writing tone (e.g., "professional", "casual", "friendly")
• Target Word Count: Desired article length
• Readability Level: "easy", "intermediate", or "advanced"
• Target Audience: Description of the intended readers
• SEO Goals: Specific SEO objectives

Example:

result = await agent.process(
    url="https://example.com/article",
    primary_keywords=["keyword1", "keyword2"],
    secondary_keywords=["keyword3"],
    tone="professional",
    target_word_count=1500,
    readability_level="intermediate",
    target_audience="Marketing professionals",
    seo_goals="Rank for primary keywords",
)

Output

Results are automatically saved to the output folder in the same directory as the script:

File Structure

• {article-title}.md - Clean article content in markdown format (title-based filename)
• {article-title}_info.txt - Article metadata, scores, and analysis

Output Location

• When run as a script: {script_directory}/output/
• When run as a module: {current_working_directory}/output/

The output folder is automatically created if it doesn't exist.

File Contents

Markdown File (.md):

• Contains only the cleaned article content
• No title, meta description, or formatting markers
• Ready for direct use or publishing

Info File (.txt):

• Article title and word count
• Meta description and character count
• Title variations
• Readability score and suggestions
• SEO score and suggestions
• Humanization details (engagement score, changes made, improvements)

Writing Style Guidelines

The humanization step applies strict writing style guidelines:

SHOULD:

• Use clear, simple language
• Be spartan and informative
• Use short, impactful sentences
• Use active voice
• Focus on practical, actionable insights
• Use "you" and "your" to directly address readers

AVOID:

• Em dashes (—)
• Metaphors and clichés
• Generalizations
• Setup language (in conclusion, etc.)
• Unnecessary adjectives and adverbs
• Hashtags, semicolons, markdown, asterisks
• A comprehensive list of banned words

Dependencies

The agent requires:

• langchain - Core LangChain framework
• langchain-openai - OpenAI integration
• langgraph - Graph-based agent orchestration
• aiohttp - Async HTTP client for web scraping
• beautifulsoup4 - HTML parsing
• python-dotenv - Environment variable loading (optional)

All dependencies are listed in requirements.txt.

Logging

The agent provides detailed logging during execution:

• Node execution status
• Content extraction progress
• Article generation parameters
• Humanization progress
• Workflow completion status

All logs are printed to the console with clear visual indicators.

Error Handling

The agent includes comprehensive error handling:

• Invalid URL validation
• Content extraction errors
• Article generation failures
• Humanization errors
• File writing errors

Errors are logged and reported clearly, allowing for easy debugging.

Article-Rewriter-AI-Agent