# Hermes Agent

> MIT self-improving agent from Nous Research — skills from experience, multi-channel gateway, any model

**Category:** AI Agents  
**Pricing:** open-source  
**URL:** https://hermes-agent.nousresearch.com  
**GitHub:** https://github.com/NousResearch/hermes-agent  
**Added:** 2026-10-12

---

## The Hook

Hermes Agent is a self-hosted agent with a closed learning loop: it writes skills from hard tasks, improves them in use, searches past sessions, and keeps a deepening model of you — reachable from a TUI or from Telegram, Discord, Slack, WhatsApp and Signal through one gateway.

## What's Good

MIT (NousResearch/hermes-agent), release v0.21.6 (2026-10-08). Install via the official curl/bash or PowerShell scripts; docs also cover MSIX on Windows. Bring any OpenAI-compatible endpoint (Nous Portal, OpenRouter, OpenAI, local, and more) and switch with `hermes model`. Full TUI with slash commands and streaming tool output; messaging channels share one gateway process with cross-platform conversation continuity. ★~253k.

## The Catch

A full personal/coding agent surface, not a drop-in Claude Code plugin — you run the gateway and own model keys, channel tokens and the memory/skill store it writes. Exact-pinned Python deps and a 3.14-oriented install path mean supply-chain and upgrade care; follow their install docs rather than improvising. Multi-channel access widens the security surface the same way OpenClaw does. Vs openclaw (freemium multi-channel assistant) and claude-code (paid Anthropic terminal agent): Hermes is the MIT self-improving host you operate yourself. oh-my-hermes is niche plugin tooling for this host, not a substitute. Stars have grown extremely fast (about 3,940 a week since the repo opened in July 2025), so judge it on a trial run, not the count.

## Verdict

The open-source agent with a real learning loop — run it if you want skills and memory you own, not another rented chat UI.