Hermes Agent

Open Source

Open-source autonomous agent platform with memory, skills, subagents, and messaging gateways

Hermes Agent is an open-source autonomous agent platform from Nous Research for running persistent agents across local and remote environments. It combines long-term memory, reusable skills, tool calling, subagents, and messaging integrations so agents can operate continuously instead of only inside a single chat or terminal session.

Features

  • Persistent memory and workspace context so agents can carry state across sessions
  • Reusable skills system for packaging task-specific instructions, scripts, and workflows
  • Subagents and multiple runtimes for delegating work across different execution environments
  • Messaging gateway integrations so agents can work through chat surfaces like Slack, Telegram, and other channels
  • Tooling for browser control, code execution, document analysis, and environment automation

So agents can…

  • Keep working over time with memory, project context, and recoverable task state
  • Hand off specialized work to subagents while a main agent coordinates the broader job
  • Operate through chat, local tooling, and remote environments as part of one continuous workflow

Related tools

Comments