v1.0.8

newton is a CLI that creates your API documentation for you with AI. This version (v1.0.8) is a stable release. The following changes were made:

• in default mode, newton checks that the input filepath exists and verifies its format before allowing the user to proceed to the next step

v1.0.7

newton is a CLI that creates your API documentation for you with AI. This version (v1.0.7) is the first stable release. The following changes were made:

• an example Next.js export was deployed at https://newton-next-export.surge.sh/ for the README
• in default mode and transmogrify mode, newton now prompts directly for the applicable file's filepath for enhanced reliability and user flexibility (e.g. having a variety of possible filenames)
• in transmogrify mode, newton checks that the input filepath exists and verifies its format before allowing the user to proceed to the next step
• fixes were made to the Next.js export template to resolve the base URL showing up as undefined, and the copy button on the code block moving while scrolling
• interface updates were made to the Next.js export template to increase the font size of the endpoint titles for better readability
• older screenshots were removed from the package build to decrease package size

v1.0.6

newton is a CLI that creates your API documentation for you with AI. This version (v1.0.6) is the first official public release. The following features are included:

• specifying a personal OpenAI API key to use newton with GPT-3.5 Turbo; saved persistently in ~/.newton
• interactive prompting to configure options in generating and exporting
• works with Express.js (Node.js) projects with an entrypoint specified in the main field of the project's package.json
• works with Flask (Python) projects with an app.py containing the routes
• offers the option to export as a Markdown file, basic HTML page (unstyled), and Next.js site; JSON is always exported by default
• allows the converting of an existing newton-generated JSON to another export option without recalling the AI

As you might've noticed, there are a few limitations that exist in parsing project directories and entrypoints. The goal is to make this more robust in future versions.