Very significant. Nearly every commit has involved the use of one or more LLMs, as evidenced by the commit trailers. I would not have started this project without it, because I do not know Rust. Even the overall direction and architecture has involved roleplay-based "rubber ducking" with LLMs [0].
I've carefully stewarded & heavily edited the Ruby code in lib/ and test/, and the documentation (RDoc and Markdown). The Rust code has been left largely to the AI, with its quality kept presumably-okay by Clippy and extensive automated tests on the Ruby side.
As for the non-library stuff ("internal" to the project), you can tell by browsing the tasks/ folder where I left the AI to its own devices [1], and where I heavily edited the Ruby code [2].
Very significant. Nearly every commit has involved the use of one or more LLMs, as evidenced by the commit trailers. I would not have started this project without it, because I do not know Rust. Even the overall direction and architecture has involved roleplay-based "rubber ducking" with LLMs [0].
I've carefully stewarded & heavily edited the Ruby code in lib/ and test/, and the documentation (RDoc and Markdown). The Rust code has been left largely to the AI, with its quality kept presumably-okay by Clippy and extensive automated tests on the Ruby side.
As for the non-library stuff ("internal" to the project), you can tell by browsing the tasks/ folder where I left the AI to its own devices [1], and where I heavily edited the Ruby code [2].
[0]: https://man.sr.ht/~kerrick/ratatui_ruby/history/ecosystem-dr...
[1]: https://git.sr.ht/~kerrick/ratatui_ruby/tree/783a08eabe2307f...
[2]: https://git.sr.ht/~kerrick/ratatui_ruby/tree/783a08eabe2307f...
Great job :-)