DEV Community: Dragos Balota

From Zero to Agent Hero: Building AI Agents with Agno + uv

Dragos Balota — Fri, 25 Jul 2025 10:52:12 +0000

Ready to build AI agents that don't just chat, but search the web, remember conversations, and dive into PDFs? Meet Agno - a lightweight Python library that's reportedly 10,000x faster than LangGraph !

What You'll Learn:

⚡ Lightning Setup with uv (10-100x faster than pip)
🤖 Basic Agent Creation with OpenAI integration
🌐 Web Search Powers using DuckDuckGo tools
🧠 Persistent Memory with SQLite storage
📚 RAG Implementation with LanceDB vector database
👥 Multi-Agent Teams that collaborate seamlessly

Quick Start:

Install Agno:

pip install agno

Create First Agno Agent

from agno.agent import Agent
from agno.models.openai import OpenAIChat

agent = Agent(
    model=OpenAIChat(id="gpt-4o"),
    description="You're a cheerful AI pal!",
    markdown=True
)

agent.print_response("Hey! What's cooking today?", stream=True)

Key Features:

🔧 Zero Configuration - Works out of the box
💾 Smart Memory - SQLite/PostgreSQL storage
🔍 RAG Ready - PDF knowledge bases with vector search
🌐 Web Integration - DuckDuckGo search capabilities
⚡ Blazing Fast - Built for speed and efficiency
🤝 Team Collaboration - Multi-agent coordination ## Real-World Example: Build a Thai cuisine expert that pulls recipes from PDFs and supplements with web research - all in under 50 lines of code!

Perfect for: Developers wanting to build production-ready AI agents without the complexity of heavyweight frameworks.

👉 🔗 Read the full guide: https://www.bitdoze.com/agno-get-start/

Running Python Scripts with uv: Zero Dependencies Management 🚀

Dragos Balota — Fri, 25 Jul 2025 08:47:21 +0000

Tired of setting up virtual environments and managing dependencies for simple Python scripts? Meet uv - the game-changing tool that lets you run Python scripts with external dependencies in seconds, not minutes!

What Makes uv Special?

✅ Zero Configuration - No virtual environments, no pip installs
✅ Lightning Fast - 10-100x faster than traditional tools
✅ Isolated Execution - Each script runs in its own environment
✅ PEP 723 Support - Embed dependencies directly in your scripts
✅ Cross-Platform - Works seamlessly on Windows, macOS, and Linux

Quick Examples

Run with external dependencies:

# API testing
uv run --with requests test_api.py

# Data analysis
uv run --with pandas --with matplotlib analyze_data.py

# Web scraping
uv run --with beautifulsoup4 --with requests scraper.py

Embed dependencies in your script (PEP 723):

# /// script
# dependencies = [
#   "requests>=2.31.0",
#   "rich>=13.0.0"
# ]
# requires-python = ">=3.10"
# ///

import requests
from rich.console import Console

# Your script code here...

Then simply run: uv run my_script.py

Real-World Use Cases

🔍 API Testing & Monitoring - Test endpoints without setup overhead 📊 Data Analysis - Process CSV files with pandas instantly 🕷️ Web Scraping - Extract and convert web content to markdown 🤖 Automation Scripts - System administration without environment hassle 🧪 Prototyping - Experiment with new libraries immediately

Sample Output

$ uv run --with requests github_api_test.py
✅ API Response Status: 200
🏢 Organization: Astral
📍 Location: United States of America
👥 Public Repos: 47
⏰ Tested at: 2025-01-08 14:30:15

Why Choose uv Over Traditional Methods?

Traditional Approach:

1. Create virtual environment (30s)
2. Activate environment (5s)
3. Install dependencies (60s+)
4. Run script
5. Clean up uv Approach:

uv run --with requests my_script.py  # 2-3 seconds total!

Advanced Features

Python Version Management : uv run --python 3.11 script.py
Executable Scripts : Use shebang #!/usr/bin/env -S uv run --script
Version Constraints : --with 'requests>=2.31.0,<3.0.0'
Multiple Dependencies : Chain multiple --with flags ## Perfect For:
DevOps Engineers - Quick automation scripts
Data Scientists - Rapid prototyping and analysis
Web Developers - API testing and content extraction
System Administrators - Monitoring and maintenance scripts
Students & Educators - Learning without setup friction ## Get Started Today!

# Install uv
curl -LsSf https://astral.sh/uv/install.sh | sh

# Run your first script
uv run --with requests my_api_test.py

Ready to revolutionize your Python scripting workflow? uv eliminates the friction between idea and execution, letting you focus on solving problems rather than managing environments.

Full Article

What's your biggest pain point with Python dependency management? Share in the comments! 👇

Bulk URL Checker with uv: Validate Website Accessibility in Python

Dragos Balota — Fri, 25 Jul 2025 08:40:10 +0000

Building a powerful URL validation tool has never been easier! With Python's uv package manager, you can create an enterprise-level URL checker that validates hundreds of websites concurrently with zero setup required.

🚀 What Makes This Special?

Zero Configuration : Run immediately with uv - no virtual environments or dependency management
Concurrent Processing : Check multiple URLs simultaneously using ThreadPoolExecutor
Smart Error Detection : Categorizes timeouts, connection errors, and HTTP status codes
Detailed Reporting : Response times, status codes, and comprehensive error analysis
File I/O : Read URLs from files and save problematic URLs for review
Cross-Platform : Works seamlessly on macOS, Windows, and Linux

🛠️ The Complete Script

Save this as url_checker.py and run with uv run url_checker.py :

#!/usr/bin/env -S uv run
# /// script
# dependencies = [
#     "requests",
# ]
# ///

import requests
from urllib.parse import urlparse
import time
from concurrent.futures import ThreadPoolExecutor, as_completed
import sys

def check_url(url, timeout=10):
    """Check if a URL is accessible and return status information."""
    if not url.startswith(('http://', 'https://')):
        url = 'https://' + url

    start_time = time.time()

    try:
        response = requests.get(url, timeout=timeout, allow_redirects=True)
        response_time = time.time() - start_time

        return {
            'url': url,
            'status': 'OK',
            'status_code': response.status_code,
            'error_type': None,
            'response_time': round(response_time, 2)
        }

    except requests.exceptions.Timeout:
        return {
            'url': url,
            'status': 'TIMEOUT',
            'status_code': None,
            'error_type': 'Connection timeout',
            'response_time': timeout
        }

    except requests.exceptions.ConnectionError as e:
        return {
            'url': url,
            'status': 'CONNECTION_ERROR',
            'status_code': None,
            'error_type': f'Connection error: {str(e)[:100]}...',
            'response_time': time.time() - start_time
        }

    except requests.exceptions.RequestException as e:
        return {
            'url': url,
            'status': 'ERROR',
            'status_code': None,
            'error_type': f'Request error: {str(e)[:100]}...',
            'response_time': time.time() - start_time
        }

def check_urls_batch(urls, timeout=10, max_workers=10):
    """Check multiple URLs concurrently."""
    results = []

    with ThreadPoolExecutor(max_workers=max_workers) as executor:
        future_to_url = {executor.submit(check_url, url, timeout): url for url in urls}

        for i, future in enumerate(as_completed(future_to_url), 1):
            result = future.result()
            results.append(result)
            print(f"Checked {i}/{len(urls)} URLs: {result['url']} - {result['status']}")

    return results

# Full script available in the complete article

📊 Perfect For:

Website Audits : Validate all external links on your site
SEO Monitoring : Check backlinks and partner sites
API Health Checks : Monitor endpoint availability
Competitor Analysis : Track competitor website status
CI/CD Integration : Automated link validation in pipelines ## 🎯 Sample Output

URL Connection Checker
==============================
Found 8 URLs to check.
Using timeout: 10 seconds

Checked 1/8 URLs: https://www.google.com - OK
Checked 2/8 URLs: https://nonexistent-site.com - CONNECTION_ERROR

==================================================
SUMMARY
==================================================
Total URLs checked: 8
Working URLs: 7
Problematic URLs: 1

WORKING URLs (7):
  ✓ https://www.google.com (Status: 200, Time: 0.15s)
  ✓ https://www.github.com (Status: 200, Time: 0.23s)

🚀 Why uv?

Instant execution : No setup, no virtual environments
Automatic dependency management : uv handles everything
Lightning fast : Faster than pip and conda
Modern Python tooling : The future of Python package management ## 💡 Advanced Features
Custom timeouts and worker counts
Automatic HTTPS protocol addition
Progress tracking with real-time updates
Categorized error reporting
Export problematic URLs to files
Support for comments in URL lists This tool has saved me countless hours in website maintenance and SEO auditing. The concurrent processing makes it incredibly fast, and the detailed error categorization helps prioritize which issues to fix first.

Try it yourself : Create a urls.txt file with your links and run uv run url_checker.py !

Want the complete script with all advanced features?

Check out the full article with detailed explanations, troubleshooting tips, and integration examples.