Rick Donato

🐍 Python Tip! Tired of typing uv run every time? Install any Python CLI as a standalone uv tool and call it directly from anywhere, with no venv and no prefix. # Before: prefix every call with `uv run` uv run mytool --help uv run mytool serve # After: install once... uv tool

NetBox Tip! Prevent direct changes to the main branch when using the NetBox Changes plugin. Enable protect_main to ensure all changes are made through branches and follow an approval workflow. Here's an example configuration: # ./configuration.py PLUGINS_CONFIG = { "netbox_changes": { "protect_main": True,

MCP has gained a huge amount of popularity. And for good reason, it's easy to use and can allow you to really harness the benefits of LLMs. So here's a quick MCP tip! Dynamically create MCP servers with FastMCP by passing a REST API's

Background A question that came up during a recent NetBox training session was: "How should authentication credentials be provided within NetBox webhooks?" For example, you may need to send a webhook from NetBox to an external system such as: * CI/CD platforms * internal APIs * configuration management systems * notification

NetBox/Ansible Tip: The netbox.netbox.nb_inventory plugin pulls data from multiple NetBox API endpoints. Enable caching and disable unused extra lookups to reduce API calls and improve inventory load times. Here’s what you need to add to your NetBox inventory plugin file: # Cache inventory data cache: True

Use uvx to run development tools without adding them to your project dependencies. uvx runs each tool in an isolated, cached environment, keeping your pyproject.toml clean and focused on actual dependencies. Here's an example: # uvx (alias for uv run tool) # Format and lint code without modifying pyproject.

Use .filter() in Config Templates to query objects during lookups and conditional logic. This exposes the Django ORM QuerySet filter directly in your templates. Here's an example: {% for iface in device.interfaces.filter(enabled=True, name__startswith="ge-") %} interface {{ iface.name }} {% endfor %}

Typically, when working with an LLM, you send your data along with a prompt and receive a response. But this pattern often introduces issues, especially with public LLMs, due to security concerns and token bloat. There is a different pattern we can use: the Schema-Driven LLM Query Pattern. With

OpenAI Tip: If you need structured output from the LLM, use structured responses and define the expected format with JSON Schema. Here’s an example: from openai import OpenAI import os client = OpenAI(api_key=os.getenv("OPENAI_API_KEY")) # Define the exact shape you want back. schema