I built a better, human like memory, for Agents

github.com

13 pointsemson2 comments

Love the idea.

Please protect your supply-chain. Python projects in the wild, especially as relates to agents, give me the willies these days with the GitHub Actions compromise that's been rolling through the OpenClaw dependencies. (Last one I saw was Telnyx.)

How does this blend with something like wedow/ticket, or beads? Those are "short term" in-repo task lists that help shape context. I've been thinking that adding a good long-term memory was the next step. I'll have to kick the tires tomorrow.

Thanks so much for the comment! Yes, I initially had LiteLLM gateway and backed it out after the security issues. Memory is becoming a crucial part of agents and there isn't a one size fits all solution - unfortunately. I found I wanted to replicate things like SOUL.md but in a way that could "evolve", so this project has "frames" which is just a collection of tagged / filtered memory blocks which can be used to identify specific concepts. It also allows the agent to use "outcomes" to calibrate it's own memory blocks, which is really powerful. Anyway I think "cognitive loops" are going to become more relevant, and will be an art. Thanks again... let me know how you get on!!

Reply on news.ycombinator.com