Show HN: PatchPal – a small, hackable Claude Code–style coding agent in Python (github.com)

PatchPal, a new lightweight coding and automation assistant inspired by Claude Code, has been announced to support both local and cloud-based large language models (LLMs). This versatile tool allows developers to build software, debug, and automate tasks with features like executable Python generation, interactive workflows, and skills management. Users can access various functionalities such as data analysis, web scraping, or API interactions through simple commands, making it an ideal tool for both learning and executing software development tasks. The significance of PatchPal lies in its open-source nature and accessibility, designed for a variety of operating systems, including Linux, MacOS, and Windows. By enabling configurations with multiple LLMs—from Anthropic’s Claude models to OpenAI’s offerings—PatchPal encourages experimentation and adaptability across different projects. Additionally, it supports LiteLLM-compatible models, with a recommendation for local deployment using vLLM technology, which delivers improved performance and reduced latency. This makes PatchPal a compelling choice for developers looking to leverage AI in their coding practices while maintaining an easily customizable architecture.