Real Estate Browsing Agent

This agent uses Pydantic AI and Microsoft's Playwright MCP to browse real estate websites and find properties that match user queries.

Supported Websites

• magicbricks.com
• 99acres.com
• housing.com
• nobroker.com

Features

• Search for properties to buy or rent
• Filter by location, price range, number of bedrooms, and other criteria
• Browse multiple real estate websites
• View property details including prices, features, and amenities
• Compare properties across different platforms

Prerequisites

• Python 3.8 or higher
• Node.js 18 or higher (for Playwright MCP)
• Anthropic API key for Claude

Installation

1. Clone this repository:

   git clone https://github.com/yourusername/real-estate-agent.git
   cd real-estate-agent
   

2. Install required Python packages:

   pip install pydantic-ai logfire python-dotenv
   

3. Install the Playwright MCP package:

   npm install -g @playwright/mcp@latest
   

4. Create a .env file in the project root by copying the example:

   cp .env.example .env
   

   Then add your actual API keys to the .env file. Never commit your actual API keys to version control.

Running the Agent

Run the agent using Python:

python agent_realestate.py

On first run, the agent will install necessary browser components. Follow the on-screen prompts to interact with the agent.

Example Queries

Here are some example queries you can try:

• "Find 2BHK apartments for rent in Bangalore within 25,000 rupees per month"
• "Show me houses for sale in Mumbai with 3 bedrooms under 1.5 crores"
• "Look for rental properties in Delhi near metro stations"
• "Find newly constructed apartments in Pune with swimming pool"
• "Search for 1BHK flats in Hyderabad on nobroker.com"

Troubleshooting

If you encounter any issues:

1. Make sure you have the latest versions of the dependencies
2. Check that your Anthropic API key is valid
3. Try running with the --headless flag if you're on a system without a display
4. Verify that you have proper internet connectivity to browse websites

Architecture Notes

• The agent uses Playwright MCP for browser automation
• Pydantic AI provides the Agent framework with MCP support
• LogFire is used for logging all interactions
• Property search templates are configured for each website

License

MIT License

Gmail Management Agent

The agent_gmail.py script provides an AI assistant that can help manage your Gmail inbox using natural language commands.

Features

• Send emails with subject, content, and recipients (including CC and BCC)
• Create draft emails without sending
• Read specific emails by ID
• Search emails using Gmail's search syntax
• List emails from inbox, sent, or custom labels
• Mark emails as read/unread
• Move emails to different labels/folders
• Delete emails
• Batch process multiple emails at once
• List all available Gmail labels
• Create, rename, and delete labels

Setup

1. Ensure you have the required dependencies:

pip install -r requirements.txt

2. Create an OAuth client ID in Google Cloud Platform:

   • Go to the Google Cloud Console
   • Create a new project or select an existing one
   • Enable the Gmail API
   • Create OAuth 2.0 credentials (Web or Desktop application)
   • Download the credentials as JSON

3. Place your credentials file (named gcp-oauth.keys.json) in either:

   • Your current working directory, or
   • ~/.gmail-mcp/ directory

4. Set your Anthropic API key in the .env file:

ANTHROPIC_API_KEY=your_api_key_here

Usage

Run the Gmail agent:

python agent_gmail.py

The first time you run the agent, it will guide you through the OAuth authentication process to connect to your Gmail account.

After authentication, you can interact with your Gmail inbox using natural language commands like:

• "Show me my unread emails"
• "Search for emails from [email protected] with attachments"
• "Send an email to [email protected] with subject Meeting Tomorrow"
• "Create a new label called Projects"
• "Move emails from [email protected] to the Projects label"

Type 'exit', 'quit', or 'bye' to end the session.

Shopify Management Agent

The agent_shopify.py script provides an AI assistant that can help manage your Shopify store using natural language commands.

Features

• View and search customers in your Shopify store
• Get customer details and order history
• Create and manage products
• Check inventory levels
• View and process orders
• Manage collections and discounts
• Get store analytics and sales data

Setup

1. Ensure you have the required dependencies:

pip install -r requirements.txt

2. Set up your Shopify credentials in the .env file:

SHOPIFY_STORE_URL=your-store.myshopify.com
SHOPIFY_ACCESS_TOKEN=your_access_token
SHOPIFY_API_VERSION=2024-04
ANTHROPIC_API_KEY=your_anthropic_api_key

3. To get your Shopify Access Token:
   • Log in to your Shopify Admin
   • Go to Apps > Develop apps
   • Create a custom app with the necessary scopes
   • Generate an admin API access token

Usage

Run the Shopify agent:

python agent_shopify.py

You can interact with your Shopify store using natural language commands like:

• "Show me a list of customers"
• "Search for customers who purchased in the last month"
• "Show me inventory for product X"
• "List all orders with status 'unfulfilled'"
• "Show me sales data for last week"

Type 'exit', 'quit', or 'bye' to end the session.

Security Best Practices

When using this repository:

1. Never commit sensitive information like API keys or tokens to git
2. Use environment variables for all secrets as shown in .env.example
3. Regenerate any tokens or API keys if they are accidentally exposed
4. Make sure .env is in your .gitignore file (already set up)
5. For the Gmail agent, store OAuth credentials outside the repository

Reddit Market Research Agent

The agent_reddit.py script provides a specialized market research assistant focused on helping entrepreneurs validate business ideas and research markets using Reddit data.

Features

• Analyze Reddit discussions for startup market validation signals
• Research competitor sentiment and positioning in various markets
• Identify customer pain points and unmet needs
• Analyze pricing strategies and customer reactions
• Track industry trends and emerging opportunities
• Map competitive landscapes and identify market players
• Categorize companies as market leaders, challengers, and new entrants
• Discover emerging startups in any market space
• Generate structured market research reports
• Target business-focused subreddits automatically
• Intelligently formulate search strategies based on query type

Setup

1. Ensure you have the required dependencies:

pip install -r requirements.txt

2. You can use the Reddit agent in two ways:

   • Using uvx (recommended): agent_reddit.py
   • Using pip installation: agent_reddit_pip.py

3. Set your Anthropic API key in the .env file:

ANTHROPIC_API_KEY=your_api_key_here

Usage

Run the Reddit Market Research agent:

python agent_reddit.py

Or if you prefer the pip installation:

python agent_reddit_pip.py

You can interact with the agent using natural language queries like:

• "Research market validation for a SaaS project management tool"
• "What pain points do small business owners mention about accounting software?"
• "Analyze competitor sentiment for Notion alternatives"
• "What pricing feedback exists for subscription-based business models?"
• "Find discussions about bootstrapped vs VC-funded startups"
• "Is there demand for a mobile app that helps small businesses manage inventory?"
• "What features do users want in collaboration software?"
• "Who are the main players in the AI design tools market?"
• "Map the competitive landscape for subscription meal kits"
• "Identify emerging startups in the no-code development space"
• "What companies are challenging Shopify in the e-commerce platform market?"

The agent will automatically:

1. Identify relevant business subreddits to search
2. Generate appropriate search queries
3. Apply the right market research framework
4. Structure the findings into actionable insights
5. Categorize market players by type (leaders, challengers, adjacent)
6. Provide specific company names whenever possible

Type 'exit', 'quit', or 'bye' to end the session.

Types of Analysis

The agent can perform several types of specialized market research:

1. Market Validation Analysis - Determine if there's demand for your business idea
2. Competitor Analysis - Research strengths and weaknesses of existing players
3. Idea Validation - Test if your concept resonates with potential users
4. Market Player Analysis - Identify and categorize companies in a market space

Security Best Practices

When using this repository:

1. Never commit sensitive information like API keys or tokens to git
2. Use environment variables for all secrets as shown in .env.example
3. Regenerate any tokens or API keys if they are accidentally exposed
4. Make sure .env is in your .gitignore file (already set up)
5. For the Gmail agent, store OAuth credentials outside the repository