Here’s how I use LLMs to help me write code

This post shares practical tips for using LLMs to write code effectively, emphasizing that it's not always easy and requires managing context, setting expectations, and thorough testing. The author suggests thinking of LLMs as over-confident, lightning-fast pair programming assistants and provides a detailed example of building a tool with Claude Code.

https://simonwillison.net/2025/Mar/11/using-llms-for-code/

Isolating Python and Jupyter using firejail

This article outlines a method for running isolated Python processes using Firejail, ensuring they can only access a specified project directory and limiting network access. The setup involves creating a virtual environment, installing packages with temporary internet access, and running Jupyter Notebooks via a Unix socket to maintain isolation.

https://bnikolic.co.uk/blog/python/firejail/2025/03/05/python-isolated

Pydoll

Pydoll is a library for automating chromium-based browsers without a WebDriver, offering realistic interactions. It supports Python's asynchronous features, enhancing performance and enabling event capturing and simultaneous web scraping.

https://github.com/thalissonvs/pydoll/

ollama-deep-researcher

Fully local web research and report writing assistant.

https://github.com/langchain-ai/ollama-deep-researcher

Pruna

Pruna is a model optimization framework built for developers, enabling you to deliver faster, more efficient models with minimal overhead.

https://github.com/PrunaAI/pruna

Flet - Desktop & Mobile apps in Python!

This video tutorial demonstrates how to use the Flet framework in Python to build cross-platform mobile and desktop apps, even without prior front-end development experience. It walks through creating a simple quiz application, explaining how to structure the UI with Flet controls and handle user interactions.

https://www.youtube.com/watch?v=4aiNStwq8oU

A Very (!) Early Play With Astral's Red Knot Static Type Checker

Astral is developing a new Python type checker, aiming to be significantly faster than mypy. The author explores building and running it, finding it promisingly fast but still a work in progress with some type checking discrepancies.

https://jurasofish.github.io/a-very-early-play-with-astrals-red-knot-static-type-checker.html

Python’s ‘shelve’ is really useful for LLM debugging

Use python’s shelve as a quick and easy, persistent key-value store for LLM workflows.

https://www.jerpint.io/blog/python-shelve-as-cache/

RA.Aid

RA.Aid (pronounced "raid") helps you develop software autonomously. It is a standalone coding agent built on LangGraph's agent-based task execution framework. The tool provides an intelligent assistant that can help with research, planning, and implementation of multi-step development tasks.

https://github.com/ai-christianson/RA.Aid

Pythonx

Pythonx runs a Python interpreter in the same OS process as your Elixir application, allowing you to evaluate Python code and conveniently convert between Python and Elixir data structures.

https://github.com/livebook-dev/pythonx

Specializing Python with E-graphs

This blog post introduces egglog, a new library that combines equality saturation and datalog, offering advantages over traditional e-graph libraries like egg. It highlights egglog's cleaner syntax, efficient execution through database techniques, and improved performance in areas like multi-pattern matching and incremental e-matching.

https://vectorfold.studio/blog/egglog

CSM

A Conversational Speech Generation Model.

https://github.com/SesameAILabs/csm

How to Make Your Python Docker Image 80% Smaller

This video provides a step-by-step guide to creating optimized and efficient Docker images for Python applications, covering topics like choosing the right base image, cleaning up dependencies, using multi-stage builds, and enhancing security. It also highlights the frustrations of working with Docker and potential pitfalls to avoid.

https://www.youtube.com/watch?v=tc713anE3UY

Virtual/AI Python Assistant - Pybot

Get ai-driven help on Python programming. You can use the assistant to debug code, generate snippets or explain python concepts.

https://www.pynerds.com/ai-assistant/

Python Developer Tooling Handbook

This handbook guides Python developers through the ecosystem of tools that enhance productivity, covering areas like packaging, linting, formatting, and dependency management. It's structured to provide tutorials, how-to guides, explanations, and references to help make sense of the complex world of Python development.

https://pydevtools.com/handbook/

Smoke test your Django admin site

This post advocates for testing Django admin changelist views, which are often overlooked despite being prone to errors due to Django's "magic-string"-heavy admin DSL. The author provides a parametric test using pytest to ensure that every admin view can render its empty state, catching potential issues early.

https://jmduke.com/posts/post/django-admin-changelist-test/

basic-memory

Basic Memory is a knowledge management system that allows you to build a persistent semantic graph from conversations with AI assistants. All knowledge is stored in standard Markdown files on your computer, giving you full control and ownership of your data.

https://github.com/basicmachines-co/basic-memory