MCPFast / Tools / MCP plugin for Claude Code: persistent memory & security
This MCP plugin for Claude Code provides persistent memory and security via HMAC, reducing token usage and securing tool calls.
View on GitHub→This MCP plugin enhances Claude Code by integrating persistent memory and robust security features. Designed for developers leveraging Claude Code, it addresses common challenges related to state management and secure interaction with external tools. By implementing a secure context management system, this plugin aims to streamline AI agent development workflows, reduce operational costs through token optimization, and ensure the integrity of tool calls.
The core functionality of this plugin is to provide Claude Code with a mechanism for persistent memory . This allows the AI to retain information across multiple interactions, avoiding the need to re-establish context with every new prompt. Furthermore, it introduces a security layer using HMAC (Hash-based Message Authentication Code) to authenticate and authorize tool calls. This ensures that only legitimate and expected actions are executed by the Claude Code agent, preventing unauthorized access or manipulation of external services.
This MCP plugin is specifically designed for AI developers who are building and deploying agents using Claude Code. It is ideal for those who require their AI agents to:
If you are working on complex AI projects involving stateful interactions and secure tool integration, this plugin offers a valuable enhancement to your Claude Code development environment.