I'm listening to the ATmosphere conference and just came across panproto: 
On the surface it’s a project about schema version control and migrations across systems like atproto, sql, protobuf, graphql, json schema, etc. which sounds a little dry until you think about what that actually means for open protocols.
One of the recurring problems in protocol ecosystems is that we treat compatibility and migration as if they’re just social coordination problems right up until they become infrastructure problems. then every client, relay, indexer, bridge, and app ends up carrying around its own private pile of assumptions.
Nostr has a lot of strengths because it evolves loosely and permissionlessly. that flexibility matters. but it also means a lot of the real schema evolution in nostr happens informally: new kinds, new tag meanings, app-specific conventions, and changing assumptions spread unevenly through the ecosystem.
Feels especially relevant for long-lived clients, relays, indexers, and definitely for anything trying to bridge nostr with atproto or other networks.
anyway, interesting project. protocol people might want to take a look
panproto
panproto | schema migration and version control
panproto parses and compares schemas, constructs migrations and lenses, transforms data, and stores schema history across supported formats.
