Supermemory: memory tool for Hermes Agent

Memory and context engine for AI agents with an MCP server, open-source plugins and a Hermes provider

What Supermemory does

Supermemory learns from conversations, extracts facts, maintains user profiles, handles knowledge updates and contradictions, and forgets expired information. The README describes hybrid search that combines RAG with memory in one query, connectors for Google Drive, Gmail, Notion, OneDrive and GitHub with real-time webhooks, and extractors for PDFs, images (OCR), videos (transcription) and code (AST-aware chunking).

There are three ways to use it. Plugins and an MCP server at https://mcp.supermemory.ai/mcp give assistants persistent memory, and the README lists plugins for Claude Code, Muse Code, Cursor, Codex, OpenCode, OpenClaw and Hermes. Developers can call a single API for memory, RAG, profiles and connectors. A local mode runs as one binary and can work fully offline with Ollama. For Hermes, the README points to the Supermemory memory provider in the NousResearch/hermes-agent repository, not to code in this repository.

Key features

Fact extraction from conversations, with handling of contradictions and expired information
Auto-maintained user profiles served in one call
Hybrid search combining RAG and memory in a single query
Connectors for Google Drive, Gmail, Notion, OneDrive and GitHub
MCP server with memory and recall tools
Local mode as a single binary, optionally fully offline with Ollama

When to use it

Giving an agent persistent memory of preferences and past projects across sessions
Building an app that needs memory, RAG and user profiles through one API
Running a self-hosted memory layer offline with Ollama

Who it is for: Developers and agent users who want a hosted or self-hosted memory layer for their AI tools, including Hermes Agent users looking for a memory provider.

How it fits with Hermes Agent

The README lists Hermes among the agents Supermemory has plugins for and links the Supermemory memory provider in the NousResearch/hermes-agent repository. This repository holds the memory engine and API rather than the Hermes-side code.

How to install Supermemory

These commands are copied from the project's README. Check the repository for the latest steps before you run them.

curl -fsSL https://supermemory.ai/install | bash

FAQ

What is Supermemory?

Supermemory is a memory and context layer for AI. It extracts facts from conversations, keeps user profiles up to date, handles contradictions and expiry, and delivers relevant context through an API, plugins or an MCP server.

Does Supermemory work with Hermes Agent?

Yes. The README lists Hermes among the supported agents and links a Supermemory memory provider that lives in the NousResearch/hermes-agent repository.

How do I install Supermemory locally?

For the single-binary local version, the README gives curl -fsSL https://supermemory.ai/install | bash. For assistants, it also offers plugins and an MCP server at https://mcp.supermemory.ai/mcp.