GitHub Repository

Git-based CLI memory harness for AI agents: structured project memory, flexible wiki, CLI first.

0 starsJavaScript

Memory Magico – CLI based memory, wiki and deterministic sprints

Name: Memory Magico – CLI based memory, wiki and deterministic sprints
Availability: InStock
Author: richsherwood

by richsherwood·Jun 21, 2026·2 points·0 comments

Visit Project View on HN

AI Analysis

●●SolidNiche GemBig Brain

Git-based agent memory makes context diffable, reviewable, and revertible.

Strengths

•Clear philosophy: cheap creation, deliberate promotion, expensive verification
•Everything versioned in git means memory changes are diffable and branchable
•Structured memory types address specific agent failure modes like context drift

Weaknesses

•Zero stars and forks suggests very early stage with unproven adoption
•Agent memory category is emerging with competing approaches (MemGPT, etc.)

Post Description

hey, I got tired of using GitHub issues and Mira to manage sprints and issue tracking so this is a tool I have been using for a while in a few of my projects to ensure that long sprints, and long bug finding sessions got triaged and properly implemented while holding the agent accountable. It works great.

Based on LMWiki where there is a raw input, and then triaged out into issues, discoveries, memories, phases and sprints. Sprints are deterministic and require validation to close. the chain of promotion becomes more and more strict and truth based.

git based, uses md and json, and jsonl format only. Built for mostly cli agentic usage, however there is a human dashboard included if wanted.