RLM Infinite Memory for Claude

Example Use Cases

• 1

  Persistent memory across multiple Claude Code sessions to retain context and insights.

• 2

  Advanced search capabilities including semantic, fuzzy, and regex search over conversation history.

• 3

  Multi-project and multi-domain organization with smart retention and archival of conversation data.

Description

RLM is an MCP server designed to extend Claude Code's memory by saving insights and conversation chunks persistently across sessions. It hooks into Claude Code's /compact event to auto-save context before it is wiped, supports semantic and fuzzy search, multi-project organization, and smart retention policies. This improves productivity by preventing repeated mistakes and loss of important context, enabling agents to recall past decisions, facts, and preferences seamlessly.