Show HN: Anneal – model-agnostic; makes LLMs remember, obey, and stay private

Just tried this out. I like how the memory isn’t a black box here.

Slick interface, nice!