Persistent AI memory workflow for Claude Code using RLM-Toolkit

A recursive language model (RLM) system with hypergraph memory for agentic coding.

A Model Context Protocol (MCP) server that provides AI agents with persistent memory and semantic file discovery.

Persistent project memory for Claude Code

Open Source Memory for LLMs based on Recursive Language Model architecture with integrated MCP server.

Production-grade Recursive Language Model (RLM) agent rebuilt on LangGraph, explicit StateGraph, checkpointed memory, sandboxed REPL, recursive sub-LLM calls, and vector search over Qdrant. Evolved from rlm-agent.

RLM-powered conversation memory plugin for Claude Code - lossless context management with intelligent retrieval

Memory augmented inference library for Recursive Language Models (RLMs), built on top of rlm.

Hierchical memory is the optimization technique applied to the 5 part memory system (cached memory, short term memory and long term memory, RLM memory and Meta Memory)

Autonomous coding agent with infinite memory (RLM) and scalable multi-agent orchestration

Production-grade Recursive Language Model (RLM) agent rebuilt on LangGraph, explicit StateGraph, checkpointed memory, sandboxed REPL, recursive sub-LLM calls, and vector search over Qdrant. Evolved from rlm-agent.

RLM Plugin v3.0 for Claude Code — Recursive Language Models with dual execution modes, git-aware incremental analysis, memory persistence, and multi-agent orchestration

Skill for memory management using faiss

Scalable conversational memory via recursive sub-agent delegation — 46% EM vs 5% truncation on LongMemEval-S, zero training

Recursive Language Model (RLM) implementation for OpenClaw - enables processing arbitrarily long prompts through recursive decomposition

External memory for AI agents - a mini RLM

OpenCode-Amauta: Enhanced AI coding agent with RLM, Memory Bank, Task Manager

Claude Code Recursive Language Model - RLM-style agent with filesystem-backed working memory

RDMD: Recursive Dynamic Memory Discovery — a hybrid DMD+RLM architecture for long-context memory retrieval with life-credits survival signal

This is an experiment to supercharge CLAUDE CODE USING RLM PRINCIPLES (Auto skills, persistent memory and infinite input context window)

RLM+Ralph is a low-cost autonomous coding loop: each iteration plans → writes code → verifies → repeats. RLM keeps context small by using external memory + file indexing, so cheap models can iterate indefinitely without context overflow, correcting mistakes until done.

🛡️ The only Open-Source RLM Memory Server with Mathematically Proven Safety. Persistent context, MCP tools & Sentinel Lattice protection for autonomous AI agents.

https://arxiv.org/html/2512.24601v1?utm_campaign=mit-just-solved-ai-s-memory-problem-with-rlms&utm_medium=Newsletter&utm_source=Generative_AI

Vrishabh — Agentic AI equity analyst for the Indian retail investor. Live TA signals, intelligence knowledge graphs, autonomous briefs, watch conditions, and persistent memory. Built on GPT-4o + RLM.

TDD-OOP Swarm Framework for Claude Code - Parallel agent orchestration with Pydantic enforcement, RLM/REPL, and persistent memory. Achieved 64/64 (100%) on Aider Polyglot Benchmark (Python + Rust).

The Ai assistant which uses context engineering to extend the amount of context if can recalls by summarising what was discussed in a 4-plane memory system and uses RLM to actively search anything from the main conversation without overloading the LLM