Hermes Agent

A self-improving AI agent with a learning loop, multi-platform messaging gateway, and a paper-themed web dashboard.

Maintained by badhope — originally created by the Nous Research team as hermes-agent. This fork preserves the original architecture while shipping a new paper-themed dashboard, multilingual UI, and a friendlier runtime layout system.

---

What is Hermes Agent?

Hermes Agent is an AI assistant that gets better the more you use it. It remembers what worked, creates reusable skills from complex tasks, and builds a deeper understanding of who you are over time.

It runs on any LLM provider — OpenAI, Anthropic, OpenRouter, local models, or your own endpoint. Switch providers with one command. No vendor lock-in.

It's not tied to your laptop. Talk to it from Telegram, Discord, Slack, WhatsApp, or Signal while it works on a cloud VM.

---

Features

Feature	Description
Terminal UI	Full TUI with multiline editing, slash-command autocomplete, conversation history, streaming tool output.
Multi-platform	Telegram, Discord, Slack, WhatsApp, Signal, and CLI — all from a single gateway process. Voice transcription, cross-platform conversation continuity.
Learning loop	Agent-curated memory with periodic nudges. Autonomous skill creation after complex tasks. Skills self-improve during use. Cross-session recall via full-text search.
Scheduled tasks	Built-in cron scheduler with delivery to any platform. Daily reports, nightly backups, weekly audits — all in natural language.
Parallel subagents	Spawn isolated subagents for parallel workstreams. Write Python scripts that call tools via RPC.
Runs anywhere	Local, Docker, SSH, Singularity, Modal, Daytona. Serverless persistence — environment hibernates when idle, wakes on demand.
Paper dashboard	Clean, warm paper-themed web UI with runtime layout controls — font size, density, corner radius, content width, auto-fit, and 16-language i18n.

---

Quick Start

Linux / macOS / WSL2 / Termux

curl -fsSL https://raw.githubusercontent.com/badhope/HermesAgent/main/scripts/install.sh | bash

Windows (PowerShell)

iex (irm https://raw.githubusercontent.com/badhope/HermesAgent/main/scripts/install.ps1)

After installation:

source ~/.bashrc    # or: source ~/.zshrc
hermes              # start chatting

---

Getting Started

hermes              # Interactive CLI
hermes model        # Choose LLM provider and model
hermes tools        # Configure enabled tools
hermes config set   # Set config values
hermes gateway      # Start messaging gateway
hermes setup        # Full setup wizard
hermes update       # Update to latest version
hermes doctor       # Diagnose issues

---

CLI vs Messaging

Action	CLI	Messaging
Start chatting	hermes	hermes gateway setup + hermes gateway start, then message the bot
New conversation	/new or /reset	/new or /reset
Change model	/model [provider:model]	/model [provider:model]
Set personality	/personality [name]	/personality [name]
Retry / undo	/retry, /undo	/retry, /undo
Browse skills	/skills or /<skill-name>	/<skill-name>
Interrupt	Ctrl+C or new message	/stop or new message

---

Dashboard

The web dashboard uses a paper-inspired theme — warm off-white canvas, deep ink text, hairline rules, and a deep-amber accent.

Runtime layout controls (System page):

• Language switcher — 16 locales
• Font size — 6 steps (XS to XXL)
• Code font size — 6 steps, independent from body text
• Display density — 3 levels
• Corner radius — 3 levels
• Content max-width — 4 presets
• Auto-fit toggle — responsive layout with CSS container queries

All settings persist to localStorage and take effect immediately without rebuild.

---

Documentation

Section	What's Covered
Quickstart	Install, setup, first conversation
CLI Usage	Commands, keybindings, personalities, sessions
Configuration	Config file, providers, models, all options
Messaging Gateway	Telegram, Discord, Slack, WhatsApp, Signal
Tools & Toolsets	40+ tools, toolset system, terminal backends
Skills System	Procedural memory, Skills Hub, creating skills
Memory	Persistent memory, user profiles
MCP Integration	Connect any MCP server for extended capabilities
Cron Scheduling	Scheduled tasks with platform delivery

---

Supported LLM Providers

• OpenAI (GPT-4o, o1, etc.)
• Anthropic (Claude 3.5 Sonnet, etc.)
• OpenRouter (200+ models)
• Nous Portal (300+ models + tool gateway)
• NVIDIA NIM (Nemotron)
• Xiaomi MiMo
• z.ai / GLM
• Kimi / Moonshot
• MiniMax
• Hugging Face
• Any OpenAI-compatible endpoint

Switch with hermes model — no code changes.

---

Contributing

git clone https://github.com/badhope/HermesAgent.git
cd HermesAgent
./setup-hermes.sh
./hermes

---

License

MIT — see LICENSE.