Persistent memory for your Nanobot agents — meet Mnemon

[chancsc]

Problem / Motivation

Lack of Persistent memory

Proposed Solution

The Nanobot integration with Mnemon is officially up! 🎉

If your Nanobot agents keep forgetting context between sessions, Mnemon fixes that.

https://github.com/mnemon-dev/mnemon#readme

Mnemon is a lightweight CLI that gives agents a shared, persistent memory store — a four-graph knowledge base with intent-aware recall, automatic deduplication, and importance decay. No API keys, no cloud, just a local SQLite database at ~/.mnemon that every agent can read and write.

Why it's useful for Nanobot:

Memory is global across all sessions and projects — one install, shared context everywhere

Works without Ollama out of the box; add nomic-embed-text for semantic/vector search

Agents make their own judgment calls on what to store — no hard rules, no embedded LLM

Give it a try and let us know how it goes.

[chengyongru]

looks like a promotion rather than a real issue

[chancsc]

I use nanobot, while looking for a solution for permenant memory and came across mnemon, so I integrate nanobot support to mnemon and share in this group, in case anyone interested in permenant memory for their nanobot

[chengyongru]

Sorry for the premature close — I initially mistook this for a promotional post. After taking a closer look at your PR (mnemon-dev/mnemon#24) and the actual integration code, this is a genuine open-source contribution. My bad.

We're trying to keep the issue tracker focused on bugs and feature requests. I'll convert this to a Discussion so others can find and discuss Mnemon integration. Thanks for building the nanobot support and sharing it here.