Some background: I've worked at companies that do and don't have data catalogs/data discovery tools, and I also spent some time trying to set existing open-source implementations up for friend's startups. But it was surprisingly hard, especially to keep them stable. Considering most people only look at a few tables, I figured there must be a way to make a more lightweight stable version. And since I (and it sounds like a lot of my friends) spend a lot of time on the command-line, I figured I'd try to build out something there...<p>Enter metaframe! It's easy to install (a single brew command), has a very quick search (leveraging fzf), allows for personal documentation, and uses a text-based backend (no kafka, janus, atlas, neo4j, sns queues), so your metadata and docs are easily accessible and even git versionable. Would love feedback/thoughts.