Hermes Agent

Hermes Agent

4.7 (567)

Hermes Agent is a self-improving AI agent built by Nous Research. It features a built-in learning loop — creating skills from experience, improving them during use, and building persistent memory across sessions. Runs on 6 backends and 15+ messaging platforms.

open-source autonomous self-improving nous-research memory skills mcp
Visit Website View Source

Hermes Agent

Hermes Agent is a self-improving AI agent built by Nous Research. It is the only agent with a built-in learning loop — creating skills from experience, improving them during use, nudging itself to persist knowledge, and building a deepening model of who you are across sessions.

Source: hermes-agent.nousresearch.com/docs

What Makes Hermes Different

It’s not a coding copilot tethered to an IDE or a chatbot wrapper around a single API. It’s an autonomous agent that gets more capable the longer it runs. It lives wherever you put it — a $5 VPS, a GPU cluster, or serverless infrastructure (Daytona, Modal) that costs nearly nothing when idle.

Core Features

Self-Improving Learning Loop

From the official docs:

  • Agent-curated memory with periodic nudges
  • Autonomous skill creation during use
  • Skill self-improvement while working
  • FTS5 cross-session recall with LLM summarization
  • Honcho dialectic user modeling

Runs Anywhere

6 terminal backends: local, Docker, SSH, Daytona, Singularity, Modal. Daytona and Modal offer serverless persistence — environment hibernates when idle, costing nearly nothing.

15+ Messaging Platforms

One gateway connects to: CLI, Telegram, Discord, Slack, WhatsApp, Signal, Matrix, Mattermost, Email, SMS, DingTalk, Feishu, WeCom, BlueBubbles, Home Assistant

Skills System

From the Skills Hub:

RegistryCount
Built-in77
Optional45
Community (LobeHub, etc.)521
Total643

16 categories including Software Dev, Creative, MLOps, Research, Translation, Productivity, Gaming, AI Agents, Apple, GitHub, Media, Security.

Built-in Tools (47 total)

  • Memory System: Persistent memory across sessions
  • Skills System: Procedural memory the agent creates and reuses
  • MCP Integration: Connect to MCP servers
  • Voice Mode: Real-time voice in CLI, Telegram, Discord, Discord VC
  • Scheduled Automations: Built-in cron with delivery to any platform
  • Delegates & Subagents: Spawn isolated subagents for parallel workstreams
  • Research-ready: Batch processing, trajectory export, RL training with Atropos

Installation

From the Installation docs:

Linux / macOS / WSL2

curl -fsSL https://raw.githubusercontent.com/NousResearch/hermes-agent/main/scripts/install.sh | bash

Android / Termux

Same installer — detects Termux automatically.

After Installation

source ~/.bashrc  # or ~/.zshrc
hermes           # Start chatting!

Prerequisites: Only Git is required. The installer automatically handles Python 3.11, Node.js v22, ripgrep, ffmpeg.

Configuration Commands

hermes model      # Choose LLM provider and model
hermes tools      # Configure enabled tools
hermes gateway setup  # Set up messaging platforms
hermes config set    # Set individual config values
hermes setup         # Full setup wizard

Messaging Gateways

Supported platforms (from Messaging docs): Telegram, Discord, Slack, WhatsApp, Signal, Matrix, Mattermost, Email, SMS, DingTalk, Feishu, WeCom, BlueBubbles, Home Assistant

About Nous Research

Built by Nous Research — the lab behind Hermes, Nomos, and Psyche models. Works with Nous Portal, OpenRouter, OpenAI, or any endpoint.

Resources