ai-memory-management

Habitus-AI is fundamentally new approach to AI Agent Memory Systems and Cognitive Harness. Dual-Cypher system runs along perpendicular graphs. Mutate edges in real time. Builds habits and dynamic personality traits. Can call tool reflectively, connect complex concepts to learned words, and send plain text outputs without needing an LLM

First-Person Agent Memory Bench. 10 Categories including fact recall, multi-hop links, temporal reasoning, fact overwrites, speaker traps, refusal, credibility, and agentic tool usage. 540K token / 60 session corpus, all in first person. Dynamic output-answer-key portion. Comprehensive report with visuals and miss breakdown.

One shared memory for every AI you use, in plain Markdown files you own.

mRAG (micro-RAG) is an agnostic backend memory encoder and retrieval system designed to segment memories into basic short form belief statements that can be retrieved and injected directly into an Agent's context using multi-head queries to inject only the most highly relevant beliefs with minimal excess.

The context quality layer for AI agents — memory that checks itself: lifecycle governance, calibrated confidence, and staged claim gates. Local-first, MCP 19 tools, DeepSeek Harness plugin.

This is the ProofRay. Your personal system to handle with memory

AI-CONTEXT turns user-controlled sources such as AI exports, repositories, notes, documents, Drive exports, and email archives into reviewed canonical memory that can be selectively disclosed to a model or agent between sessions. AI-CONTEXT works by itself. QSOL-SUBSTRATE is optional.

Mega Brain MCP is a local-first knowledge control plane for software projects. It exposes six stable MCP tools while keeping AgentMemory, Code Review Graph, and Git behind versioned private adapters.

A local, offline coding agent built from scratch on Ollama + LangGraph — real tool-calling (not prompted JSON), persistent SQLite + ChromaDB memory, CPU-only, zero API cost. Built to understand how agentic tools like Cursor and Kiro actually work under the hood.

Optional skills for Hermes Agent — memory discipline, audit, and self-improvement loop. Decision trees, source-based docs, and tools for rigorous LLM memory management.

a small project management and memory extension for pi

Context hub