🤖 AI Agent - Developer Tools Research Assistant

An intelligent AI agent that researches and analyzes developer tools, technologies, and platforms to provide comprehensive recommendations. Built with LangGraph, DeepSeek AI, and Firecrawl for web scraping and analysis.

---

📸 Demo Screenshots

🚀 Tool Discovery & Analysis

AI Agent discovering and analyzing developer tools

🔍 Detailed Company Research

Comprehensive analysis of tool features, pricing, and tech stacks

💡 Intelligent Recommendations

AI-generated actionable recommendations for developers

---

🚀 Features

🔍 Discovery	🧠 Intelligence	📊 Analysis	💡 Recommendations

✨ Key Capabilities

• 🔍 Intelligent Tool Discovery: Automatically finds and extracts relevant developer tools from articles and web content
• 📊 Comprehensive Analysis: Analyzes pricing models, tech stacks, API availability, and integration capabilities
• 🤖 DeepSeek AI Integration: Uses DeepSeek AI through OpenRouter for intelligent content analysis
• 🕷️ Web Scraping: Leverages Firecrawl for efficient web content extraction
• 💡 Structured Recommendations: Provides detailed, actionable recommendations for developer tools

🏗️ Architecture

🔄 LangGraph Workflow Pipeline

graph LR
    A[🔍 User Query] --> B[📰 Tool Extraction]
    B --> C[🏢 Company Research]
    C --> D[💡 Recommendation Generation]
    D --> E[📊 Final Results]
    
    style A fill:#e1f5fe
    style B fill:#f3e5f5
    style C fill:#e8f5e8
    style D fill:#fff3e0
    style E fill:#fce4ec

Loading

🎯 Three-Stage Process

Stage	🔍 Tool Extraction	🏢 Company Research	💡 Recommendation Generation
Input	User query	Extracted tool names	Analyzed company data
Process	Web scraping + AI extraction	Website analysis + AI parsing	Data synthesis + AI recommendations
Output	List of relevant tools	Detailed company profiles	Actionable recommendations

The project uses a sophisticated LangGraph workflow with three main stages:

1. 🔍 Tool Extraction: Searches for articles and extracts relevant tool names using AI
2. 🏢 Company Research: Analyzes individual tools and their features in detail
3. 💡 Recommendation Generation: Provides comprehensive, actionable recommendations

📋 Prerequisites

🛠️ Required Tools

• 🐍 Python 3.12+ - Modern Python with latest features
• 📦 uv - Fast Python package manager
• 🔑 Firecrawl API Key - For web scraping capabilities
• 🤖 OpenRouter API Key - For DeepSeek AI access (optional)

🛠️ Installation

🚀 Quick Start Guide

📥 1. Clone the Repository

git clone <your-repository-url>
cd advance-agent

📦 2. Install Dependencies

The project uses uv for fast dependency management:

uv sync

🔧 3. Set Up Environment Variables

Create a .env file in the project root:

# Firecrawl API Key (required for web scraping)
FIRECRAWL_API_KEY=your_firecrawl_api_key_here

# Optional: OpenRouter API Key (if you want to use your own)
OPENROUTER_API_KEY=your_openrouter_api_key_here

🔑 4. Get API Keys

🔥 Firecrawl API Key	🤖 OpenRouter API Key

🔥 Firecrawl API Key (Required)

1. 🌐 Visit Firecrawl
2. 📝 Sign up for an account
3. 🎯 Navigate to your dashboard
4. 🔑 Copy your API key

🤖 OpenRouter API Key (Optional)

1. 🌐 Visit OpenRouter
2. 💳 Sign up and add credits to your account
3. 🎯 Navigate to your dashboard
4. 🔑 Copy your API key

🚀 Usage

🎯 Quick Start

📝 Basic Usage

from src.workflow import Workflow

# 🚀 Initialize the workflow
workflow = Workflow()

# 🔍 Run research on a specific query
result = workflow.run("database management tools")

# 📊 Access the results
print(f"Query: {result.query}")
print(f"Extracted Tools: {result.extracted_tools}")
print(f"Companies Analyzed: {len(result.companies)}")
print(f"Analysis: {result.analysis}")

🔍 Example Queries

🗄️ Databases	☁️ Cloud	📊 Monitoring	🔐 Security

# 🗄️ Research database tools
result = workflow.run("PostgreSQL alternatives")

# ☁️ Research deployment platforms  
result = workflow.run("cloud deployment platforms")

# 📊 Research monitoring tools
result = workflow.run("application monitoring tools")

# 🔐 Research authentication services
result = workflow.run("authentication providers")

📁 Project Structure

🗂️ File Organization

📦 advance-agent/
├── 📁 src/
│   ├── 📄 __init__.py
│   ├── 🚀 workflow.py          # Main workflow implementation
│   ├── 📊 models.py            # Pydantic data models
│   ├── 🕷️ firecrawl.py         # Firecrawl service wrapper
│   └── 🤖 prompts.py           # AI prompts for analysis
├── ⚙️ pyproject.toml           # Project configuration
├── 🔒 uv.lock                  # Dependency lock file
├── 📸 1st.jpg                  # Demo screenshot 1
├── 📸 2nd.jpg                  # Demo screenshot 2
├── 📸 3rd.jpg                  # Demo screenshot 3
└── 📖 README.md               # This file

File	Purpose	Status
	Core workflow implementation	✅
	Pydantic data structures	✅
	Firecrawl service wrapper	✅
	AI prompt templates	✅

🔧 Configuration

DeepSeek AI Configuration

The project is configured to use DeepSeek AI through OpenRouter. The configuration is in src/workflow.py:

self.client = OpenAI(
    base_url="https://openrouter.ai/api/v1",
    api_key="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",
)

Model Used: deepseek/deepseek-chat-v3-0324:free

Customizing the Model

To use a different model or your own OpenRouter API key:

1. Update the API key in src/workflow.py
2. Change the model name in the _create_deepseek_llm method

🔍 How It Works

1. Tool Extraction Phase

def _extract_tools(self, state: ResearchState):
    # Search for articles about the query
    articles_query = f"{state.query} tools comparison best alternatives"
    search_results = self.firecrawl.search_companies(articles_query, num_results=5)
    
    # Scrape content from found articles
    # Extract tool names using DeepSeek AI
    # Return list of extracted tools

What it does:

• Searches for articles related to the query
• Scrapes content from found URLs
• Uses AI to extract specific tool names from the content
• Returns a list of relevant tools

2. Company Research Phase

def _get_company_info(self, state: ResearchState):
    # For each extracted tool
    for tool_name in tool_names:
        # Search for the tool's official website
        # Scrape the website content
        # Analyze the content using AI
        # Extract pricing, tech stack, features, etc.

What it does:

• Takes the extracted tool names
• Searches for each tool's official website
• Scrapes detailed information from the websites
• Uses AI to analyze pricing, tech stack, API availability, etc.

3. Recommendation Generation Phase

def _analyze_step(self, state: ResearchState):
    # Combine all analyzed company data
    # Generate comprehensive recommendations
    # Provide actionable insights

What it does:

• Combines all the analyzed company information
• Uses AI to generate comprehensive recommendations
• Provides actionable insights and comparisons

📊 Data Models

CompanyAnalysis

class CompanyAnalysis(BaseModel):
    pricing_model: str = "Unknown"  # Free, Freemium, Paid, Enterprise
    is_open_source: Optional[bool] = None
    tech_stack: List[str] = []
    description: str = ""
    api_available: Optional[bool] = None
    language_support: List[str] = []
    integration_capabilities: List[str] = []

CompanyInfo

class CompanyInfo(BaseModel):
    name: str
    description: str
    website: str
    pricing_model: Optional[str] = None
    is_open_source: Optional[bool] = None
    tech_stack: List[str] = []
    competitors: List[str] = []
    api_available: Optional[bool] = None
    language_support: List[str] = []
    integration_capabilities: List[str] = []
    developer_experience_rating: Optional[str] = None

ResearchState

class ResearchState(BaseModel):
    query: str
    extracted_tools: List[str] = []
    companies: List[CompanyInfo] = []
    search_results: List[Dict[str, Any]] = []
    analysis: Optional[str] = None

🤖 AI Prompts

The project uses carefully crafted prompts for different analysis tasks:

Tool Extraction Prompt

Extracts specific tool names from articles and content.

Tool Analysis Prompt

Analyzes company websites to extract:

• Pricing model
• Tech stack
• API availability
• Language support
• Integration capabilities

Recommendation Prompt

Generates actionable recommendations based on all analyzed data.

🔧 Troubleshooting

Common Issues

1. Firecrawl API Key Error

ValueError: FIRECRAWL_API_KEY is not set in the environment variables

Solution: Make sure you have set the FIRECRAWL_API_KEY in your .env file.

2. DeepSeek API Error

openai.AuthenticationError: Invalid API key

Solution: The project uses a pre-configured API key. If you want to use your own, update it in src/workflow.py.

3. Import Errors

ModuleNotFoundError: No module named 'openai'

Solution: Run uv sync to install all dependencies.

4. Search Results Format Issues

AttributeError: 'tuple' object has no attribute 'get'

Solution: The code handles multiple response formats from Firecrawl. This error should be resolved in the current version.

Debug Mode

To enable debug logging, you can add print statements in the workflow methods or modify the logging level.

🧪 Testing

Test the Installation

# Test that all imports work
uv run python -c "from src.workflow import Workflow; print('✓ All imports successful')"

Test the Workflow

from src.workflow import Workflow

# Test with a simple query
workflow = Workflow()
result = workflow.run("database tools")
print(f"Found {len(result.extracted_tools)} tools")
print(f"Analyzed {len(result.companies)} companies")

📈 Performance

⚡ Speed Metrics

Stage	⏱️ Time	🚀 Status
Tool Extraction	~30-60 seconds
Company Analysis	~2-5 minutes per tool
Recommendation Generation	~10-30 seconds

🎯 Total Performance

⚡ Total time: 3-10 minutes depending on the number of tools found and analyzed.

🔒 Security

🛡️ Security Features

🔐 API Security	🚫 Data Protection	🌐 Web Ethics	⚡ Rate Limiting

• 🔐 API Security: API keys are stored in environment variables
• 🚫 Data Protection: No sensitive data is logged or stored
• 🌐 Web Ethics: All web scraping follows robots.txt guidelines
• ⚡ Rate Limiting: Rate limiting is implemented for API calls

🤝 Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add some amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

📝 License

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

🙏 Acknowledgments

• LangGraph for workflow orchestration
• DeepSeek AI for intelligent analysis
• Firecrawl for web scraping capabilities
• OpenRouter for AI model access

📞 Support

If you encounter any issues or have questions:

1. Check the Troubleshooting section
2. Search existing Issues
3. Create a new issue with detailed information

🔄 Updates

Stay updated with the latest changes:

git pull origin main
uv sync

---

🎉 Ready to Get Started?

git clone <your-repo-url>
cd advance-agent
uv sync
python example.py

🌟 Star This Repository

If this project helps you, please give it a ⭐️!

📞 Need Help?

---

Made with ❤️ by Surya Prakash Subudhiray

Happy coding! 🚀