Measure a context plugin on your own repository before you install it

Developer ArcticFox2029 has released version 1.23.1 of 'chamnan', an open-source tool designed to optimize context management for AI agents. The release introduces a web-based measurement tool that allows developers to analyze how a repository would be processed by the plugin without requiring server uploads or API keys. By running the tool locally in the browser, users can see exactly what data is injected into an AI agent's context and how the redactor handles their code. The author emphasizes that the tool provides a 'bound' rather than simple compression, ensuring consistent token usage regardless of repository size. The update also addresses several security and stability defects, including issues with path handling, Unicode sanitization, and concurrency. Built with standard library dependencies, the project aims to provide a transparent, secure way to manage AI context windows across multiple programming languages and operating systems.