i always had this question when do i know that i have enough information that a AI can follow my readme. Well this is as well one of the intersting use case of readme2demo to validate AI readiness of document.
Curious, what other use case do you see for the readme2demo and what you would like to see in future?
Love the idea of vetting tutorials before they go live. One thing I'd want as a user is a small badge or visual indicator showing the last sandbox run time and whether the demo actually completed successfully, so I can quickly tell which tutorials are still trustworthy versus stale.
@zmravbal Thanks, that's exactly the direction we want to take it. The nice part is the data already exists: every demo runs in a fresh container before it's published, and nothing ships unless that run passes. So the "last run time" and "did it actually complete" are already captured by the pipeline, surfacing them as a badge is mostly a presentation layer on top of what we're already recording, so we will take feedback and improvise.
This weekend we lauch version 0.5.0. the version extended the support to openai and gemini api key with Openhands and several bugs and patches are fixed. would love to get your feedbacks.