Openflows

Signal

BettaFish presents an open source memory plugin architecture for AI tools and local model workflows.

Context

Its structure emphasizes adapters and data connectors, treating memory as a composable layer instead of a fixed built-in feature.

Relevance

For Openflows, this sharpens a key transition: once local inference is available, memory governance becomes the next agency layer. Retrieval scope and retention control become design decisions.

Current State

Early but technically directional.

Open Questions

• Which memory interfaces are transparent enough for non-expert operators?
• How should retention and deletion be governed across local contexts?
• What tradeoff between convenience and inspectability is acceptable?

Connections

• Linked to local-inference-baseline as an extension path.

Updates

2026-03-15: Repository metrics visible on the page indicate 38.7k stars and 7.2k forks, marking a significant adoption shift from the 'early' status previously recorded. This scale validates memory governance as a viable design decision for Openflows.