I kept running into the same issue.every new session, Cursor had zero memory of decisions I’d made 20 commits ago, so I was manually re-pasting context or re-explaining architecture choices. Found a big HN thread where a ton of other devs described the exact same pain point with LLM coding in established codebases, so I figured it was worth solving properly instead of just living with it.
Built a git post-commit hook that automatically sends your diff to a small n8n workflow, which updates a CONTEXT.md file and writes straight into .cursor/rules/context.mdc.So Cursor auto-loads it, no manual @-tagging needed. Runs on every commit, fully self-hosted.
Demo (same question asked before/after): https://www.youtube.com/watch?v=eUNE-IExRIc