Function validate_urls_tool

backend/app/agents/tools/scraping_tools.py:251–289

Validate a list of URLs to check if they are accessible. This tool checks if URLs are valid and accessible, helping to filter out broken links before scraping. Args: urls: List of URLs to validate Returns: Dictionary with validation results for

(urls: List[str])

@tool("validate_urls")
async def validate_urls_tool(urls: List[str]) -> Dict[str, Any]:
    """
    Validate a list of URLs to check if they are accessible.
    
    This tool checks if URLs are valid and accessible, helping to
    filter out broken links before scraping.
    
    Args:
        urls: List of URLs to validate
        
    Returns:
        Dictionary with validation results for each URL
    """
    import aiohttp
    
    results = []
    
    async with aiohttp.ClientSession() as session:
        for url in urls:
            try:
                async with session.head(url, timeout=5, allow_redirects=True) as response:
                    results.append({
                        "url": url,
                        "valid": response.status < 400,
                        "status_code": response.status,
                        "final_url": str(response.url)
                    })
            except Exception as e:
                results.append({
                    "url": url,
                    "valid": False,
                    "error": str(e)
                })
                
    return {
        "urls_checked": len(urls),
        "valid_urls": sum(1 for r in results if r.get("valid", False)),
        "results": results
    }


# Export all tools