Lluminy is an AI documentation automation tool for Python projects. Its core purpose is to automatically generate docstrings, helping developers reduce the time spent writing documentation and focus on feature development.
Main capabilities:
- Automatically generates docstrings for Python codebases, completing documentation generation for an entire project within minutes.
- The algorithm only operates on docstrings and does not modify the original code, avoiding changes to code logic caused by large language model hallucinations.
- Supports processing multiple files or even an entire project.
- You can choose to generate documentation for all functions, or only fill in documentation for functions missing docstrings.
- Integrates with GitHub: you can connect a GitHub account and select repositories (supporting personal and organization repositories), choose specific Python files, directories, or an entire branch, and after generating results create a pull request, with no need to manually copy and paste code.
- A GitHub Bot is under development and will continuously scan pull requests and provide code comment suggestions in the future.
How to use:
- Connect a GitHub account.
- Select the repository for which documentation needs to be generated.
- Select the Python files, directories, or branch to document.
- Customize the output: generate for all functions or only for functions missing docstrings.
- Click generate, and Lluminy generates docstrings for the entire codebase and creates a pull request to merge the results.
Use cases:
- Small hobby projects: provides a limited amount of AI token usage, docstring generation is powered by LLama3.3, and supports documenting all functions or filling in missing ones.
- Teams and production codebases: provides a monthly limit of 10 million AI tokens, as well as priority support.
Limitations:
- Currently mainly aimed at Python projects.
- The GitHub Bot has not yet been released and is under development.
- Specific pricing, free quota details, platform support scope, etc. are not yet provided.