in no particular order...
- [/] network robustitude
- fix network mocks for tests
- failure handling for relays: bad payloads, disconnects, incompatible nipsets
- [/] cache for profiles + messages
- localstorage instead of idb? (simpler, maps directly to files + other k/v backends)
- add edit log, push that for sync
- [/] local filter matching
- batch filtering
- views/projections
- deno relay
- dev + build tooling
- library is "vanilla" esm
- relay "workbench" (build-a-relay w/extensions and config in code)
- metrics endpoint
- internal/service
- client-submitted