Claude API Connector

A flexible Python framework for connecting any API to Claude, enabling seamless integration and data exchange between Claude and external services.

Features

• 🔌 Universal API Connector: Connect to any REST API with minimal configuration
• 🤖 Claude Integration: Built-in Claude API client with conversation management
• 🛡️ Security First: Secure API key management and request validation
• 📊 JSON/Text Support: Primary support for JSON and text data with CSV/XML utilities
• ⚡ Async Support: High-performance async operations
• 🔄 Error Handling: Comprehensive error handling and retry logic
• 📖 Working Examples: Ready-to-use examples for popular APIs

Installation

From Source

git clone https://github.com/theRealDanB/claude-api-connector.git
cd claude-api-connector
pip install -r requirements.txt
pip install -e .

Environment Setup

Create a .env file:

ANTHROPIC_API_KEY=your_anthropic_api_key_here
DEFAULT_TIMEOUT=30
MAX_RETRIES=3
LOG_LEVEL=INFO

Quick Start

import asyncio
from claude_api_connector import ClaudeConnector, APIConfig

async def main():
    # Configure your external API
    api_config = APIConfig(
        base_url="https://api.example.com",
        headers={"Authorization": "Bearer YOUR_API_KEY"},
        timeout=30
    )

    # Initialize Claude connector
    connector = ClaudeConnector(
        anthropic_api_key="your-claude-api-key",
        api_config=api_config
    )

    # Connect and query
    result = await connector.query_with_api_data(
        prompt="Analyze this weather data",
        api_endpoint="/weather/current",
        api_params={"city": "San Francisco"}
    )

    print(result["response"])
    await connector.close()

# Run the example
asyncio.run(main())

Working Examples

Test with Public API (No Keys Required)

import asyncio
from claude_api_connector import ClaudeConnector, APIConfig

async def test_basic():
    # JSONPlaceholder - free public API
    config = APIConfig(base_url="https://jsonplaceholder.typicode.com")
    connector = ClaudeConnector(api_config=config)
    
    result = await connector.query_with_api_data(
        prompt="Tell me about this user's profile",
        api_endpoint="/users/1"
    )
    
    print(result["response"])
    await connector.close()

asyncio.run(test_basic())

Weather API Integration

from claude_api_connector import ClaudeConnector, APIConfig
from claude_api_connector.core.config import AuthType

async def weather_example():
    weather_config = APIConfig(
        base_url="https://api.openweathermap.org/data/2.5",
        auth_type=AuthType.API_KEY,
        auth_param="appid",
        api_key="your_weather_api_key"
    )
    
    connector = ClaudeConnector(api_config=weather_config)
    
    result = await connector.query_with_api_data(
        prompt="What's the weather like? Any recommendations?",
        api_endpoint="/weather",
        api_params={"q": "London", "units": "metric"}
    )
    
    print(result["response"])
    await connector.close()

Core Features

Batch Processing

endpoints = [
    {"endpoint": "/users", "params": {"active": True}},
    {"endpoint": "/orders", "params": {"status": "completed"}},
    {"endpoint": "/products", "params": {"category": "electronics"}}
]

result = await connector.batch_process(
    endpoints=endpoints,
    analysis_prompt="Provide a comprehensive business overview"
)

print(result["analysis"])

Conversation Memory

# First query - Claude remembers this context
result1 = await connector.stream_conversation(
    prompt="Analyze this sales data",
    api_endpoint="/sales/monthly"
)

# Follow-up query - Claude maintains context
result2 = await connector.stream_conversation(
    prompt="Now compare it to last year",
    api_endpoint="/sales/yearly"  
)

Custom Data Processing

def custom_processor(api_response):
    # Your custom data transformation logic
    processed_data = transform_data(api_response)
    return processed_data

connector.set_data_processor(custom_processor)

Supported APIs

This connector works with any REST API, including:

• Weather: OpenWeatherMap, AccuWeather
• Development: GitHub, GitLab, Jira
• Social Media: Twitter, Reddit, LinkedIn
• E-commerce: Shopify, WooCommerce, Stripe
• Cloud Services: AWS, Google Cloud, Azure
• And many more...

Testing

Run the included test suite:

# Set your Claude API key
export ANTHROPIC_API_KEY="your_key_here"

# Run tests (uses public APIs, no additional keys needed)
python main.py

# Run specific tests  
pytest tests/

Error Handling

from claude_api_connector.core.exceptions import (
    APIConnectionError, 
    ClaudeAPIError
)

try:
    result = await connector.query_with_api_data(
        prompt="Analyze this data",
        api_endpoint="/data"
    )
except APIConnectionError as e:
    print(f"API connection failed: {e}")
except ClaudeAPIError as e:
    print(f"Claude API error: {e}")

Important Notes

Claude API Costs

• Each query to Claude costs based on input/output tokens
• Large datasets will increase costs
• Consider data size limits and preprocessing

Rate Limits

• Respects both external API and Claude rate limits
• Built-in retry logic with exponential backoff
• Configurable concurrent request limits

Data Size Considerations

• Large API responses are automatically truncated for Claude
• Implement custom processors for data summarization
• Consider batch processing for multiple small requests vs. large single requests

Development

# Development setup
git clone https://github.com/theRealDanB/claude-api-connector.git
cd claude-api-connector
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate
pip install -r requirements-dev.txt
pip install -e .

# Run tests
pytest tests/

# Format code
black claude_api_connector/ tests/ examples/

Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Make your changes and add tests
4. Ensure tests pass (pytest)
5. Format code (black .)
6. Commit your changes (git commit -m 'Add amazing feature')
7. Push to the branch (git push origin feature/amazing-feature)
8. Open a Pull Request

Troubleshooting

Common Issues

1. "No module named claude_api_connector"

   pip install -e .

2. "ANTHROPIC_API_KEY not found"

   export ANTHROPIC_API_KEY="your_key_here"
   # or create .env file

3. API Authentication Errors

   • Verify your API keys are correct
   • Check the API documentation for auth requirements
   • Ensure proper AuthType (BEARER, API_KEY, etc.)

4. Rate Limit Errors

   • Increase delays between requests
   • Reduce batch_process concurrent limit
   • Check API provider rate limits

License

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

Support

• 🐛 Issue Tracker
• 💬 Discussions

---

Built with ❤️ for seamless AI-API integration