A while back I was asking for a safe way to use Claude Code with sensitive, proprietary data and Kedro. I came up with a Docker set up to let that happen, without using synthetic data. If you have similar challenges check this out. Feedback is welcome! gia-guar/blinded-claude: A two-container Docker starter that lets Claude Code build, run, and debug Kedro pipelines without ever seeing the data

This looks very nice! Thank you for sharing :)