MCPFast / Tools / Remem: Local persistent memory for Claude Code & Codex CLI

GitHubTool★★★★☆

Remem: Local persistent memory for Claude Code & Codex CLI

A Rust CLI tool for local persistent memory, integrating hooks, MCP server, and SQLite/SQLCipher for auditable coding work recall.

View on GitHub

Remem: Local Persistent Memory for Claude Code & Codex CLI

Remem is a Rust CLI tool designed to provide local persistent memory for your AI coding workflows, specifically targeting Claude Code and Codex CLI. It offers a robust solution for storing, retrieving, and auditing your coding sessions, ensuring that valuable insights and code snippets are never lost. By integrating with MCP servers and leveraging SQLite/SQLCipher, Remem creates an auditable log of your development process, enhancing reproducibility and collaboration.

What Remem Does

Remem acts as a local memory store for your AI coding interactions. It captures the prompts you send to AI models like Claude Code and Codex CLI, along with their generated responses. This data is then persistently stored on your local machine, allowing you to revisit past coding sessions, analyze your AI's performance, and reconstruct specific code generation steps. The tool also supports integration with MCP servers, facilitating a more centralized or distributed memory management approach if needed.

Key Features

Who Remem is For

Remem is an essential tool for AI developers, researchers, and anyone heavily involved in using AI for code generation. If you frequently use Claude Code or Codex CLI and need a reliable way to track your progress, debug issues by revisiting past AI outputs, or maintain an auditable record of your AI-assisted development, Remem is designed for you. It's particularly beneficial for teams that require reproducible development environments and transparent AI usage logs.