mnemosyne-os is a zero-dependency Python package that provides local memory and lexical retrieval for AI agents. It is designed for developers who need token-efficient agent memory without relying on hosted infrastructure.
mnemosyne-os sits in PulseGate's Memory & skills category. It focuses on giving AI agents local, token-efficient memory and retrieval without external services. mnemosyne-os is an open-source project aimed at AI developers building agent systems. The project is open source (MIT). It runs on the command line, and it can be self-hosted.
Behind mnemosyne-os is FrankHu-HK, and it first shipped in 2026. The project is developed in the open on GitHub with 45 commits in the last 90 days. Among its 6 catalogued features are lexical cache, agent memory, and local retrieval.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do