<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Memory-System on Noureddine RAMDI</title><link>https://ramdi.fr/tags/memory-system/</link><description>Recent content in Memory-System on Noureddine RAMDI</description><generator>Hugo</generator><language>en</language><lastBuildDate>Sat, 23 May 2026 20:41:27 +0000</lastBuildDate><atom:link href="https://ramdi.fr/tags/memory-system/index.xml" rel="self" type="application/rss+xml"/><item><title>cass-memory: a TypeScript cognitive memory system with confidence decay for AI coding agents</title><link>https://ramdi.fr/github-stars/cass-memory-a-typescript-cognitive-memory-system-with-confidence-decay-for-ai-coding-agents/</link><pubDate>Mon, 04 May 2026 10:23:02 +0000</pubDate><guid>https://ramdi.fr/github-stars/cass-memory-a-typescript-cognitive-memory-system-with-confidence-decay-for-ai-coding-agents/</guid><description>cass-memory is a TypeScript CLI implementing a three-layer cognitive architecture for AI coding agents, featuring a novel confidence decay system to maintain a reliable, evolving knowledge base.</description></item><item><title>Meta-Harness: evolving the scaffolding around large language models for optimized task performance</title><link>https://ramdi.fr/github-stars/meta-harness-evolving-the-scaffolding-around-large-language-models-for-optimized-task-performance/</link><pubDate>Mon, 04 May 2026 10:23:02 +0000</pubDate><guid>https://ramdi.fr/github-stars/meta-harness-evolving-the-scaffolding-around-large-language-models-for-optimized-task-performance/</guid><description>Meta-Harness from Stanford IRIS Lab automates the search for optimal harness configurations around LLMs, evolving memory, retrieval, and context systems for better task-specific performance.</description></item><item><title>MemKraft: local-first memory for AI agents with empirical self-improvement</title><link>https://ramdi.fr/github-stars/memkraft-local-first-memory-for-ai-agents-with-empirical-self-improvement/</link><pubDate>Mon, 04 May 2026 10:23:01 +0000</pubDate><guid>https://ramdi.fr/github-stars/memkraft-local-first-memory-for-ai-agents-with-empirical-self-improvement/</guid><description>MemKraft is a zero-dependency local-first memory system storing AI agent knowledge as Markdown, featuring bitemporal tracking, hybrid search, and a prompt self-improvement loop.</description></item><item><title>Mercury Agent: A TypeScript AI assistant with persistent "Second Brain" memory and permission-hardened safety</title><link>https://ramdi.fr/github-stars/mercury-agent-a-typescript-ai-assistant-with-persistent-second-brain-memory-and-permission-hardened-safety/</link><pubDate>Sat, 25 Apr 2026 19:24:00 +0000</pubDate><guid>https://ramdi.fr/github-stars/mercury-agent-a-typescript-ai-assistant-with-persistent-second-brain-memory-and-permission-hardened-safety/</guid><description>Mercury Agent is a TypeScript AI assistant with a persistent SQLite-based memory system, permission-hardened tools, and extensible skills, designed for safe, adaptive 24/7 operation via CLI or Telegram.</description></item></channel></rss>