Limitations

effect-cqrs is a trial port, not a maintained library. These are the gaps that matter before using it for anything real.

• New events stay in their process. They wake view models and subscribers in the same process only. There is no transport like the Moleculer broker, so dispatch is the entry point to wire into your own HTTP or RPC layer.
• SQLite, PostgreSQL and memory only. There are no adapters for MongoDB or MySQL.
• No persistent snapshots. Recent state is only kept in memory, so after a restart each aggregate is replayed the first time it is used, at a cost that grows with its length.
• No schema migration. A stored event that no longer decodes fails with EventDecodeError, so evolve event schemas compatibly.
• No code generator. There is no equivalent of cqrs generate.
• Slow handlers don’t slow writers. There is no backpressure, so a view model with a slow handler falls behind and catches up when writes stop.
• One thread. Throughput doesn’t grow with concurrency, and synchronous SQLite calls block the event loop while they run.