An open-source, modular MCP-aligned server/framework for building custom memory agents with hybrid vector and symbolic memory. Ideal for developers seeking full control over their MCP server setup and agent logic.
No content provided
Loading more......
A lightweight, modular MCP-compatible memory and agent protocol server designed for privacy-first, on-device AI agents. Supports local and cloud-based LLMs with persistent, session-aware memory. Also integrates real-time, up-to-date documentation directly into your coding environment, providing accurate and context-aware API information to streamline development workflows and reduce errors.
Stagehand is an open-source MCP (Model Context Protocol) server from Browserbase that enables AI agents to interact with the web, simulating browser actions such as clicking links and scraping text. It is designed to empower AIs with the ability to surf the web programmatically, making it directly relevant as an MCP server solution.
Server enabling the batching of multiple MCP tool calls into a single request, reducing token usage and network overhead for AI agents.
mcp-agent allows building effective AI agents with MCP servers using simple, composable patterns for enhanced agent functionality.
An MCP server that uses Mem0 to efficiently manage coding preferences for AI agents. It enables storing, retrieving, and searching coding snippets and preferences with context, supporting enhanced memory and reasoning for agents.
Minestom is a developer-centric, modular MCP server written in Java. It provides support for blocks, entities, inventory management, and commands, making it ideal for building custom, extensible Minecraft server environments.