All directories
AgentEngineeringOperations

Hermes Agent

nousresearch/hermes-agent

Self-improving agent from Nous Research — learns skills from experience, remembers you across sessions, runs on a VPS or serverless.

00

GitHub repository

https://github.com/nousresearch/hermes-agent
01

Overview

Hermes is an agent with a learning loop: it creates skills from hard tasks, improves them while using them, searches past chats, and can live on Telegram/Discord/Slack/WhatsApp — not only your laptop. Repo (copy/paste): https://github.com/nousresearch/hermes-agent

SaaSOther

Work with LimeDock

Using this skill? LimeDock can wire it into a durable automation you own.

Agents are a starting point. LimeDock turns the workflow into production automation your SaaS team owns — not another prompt library.

We sell owned automations for SaaS teams — live workflows that plug into Slack, CRM, and your internal platform — not just a skill list.

02

Link

https://github.com/nousresearch/hermes-agent
03

Installation guide

Not installed into LimeDock.

1. Copy: https://github.com/nousresearch/hermes-agent 2. Install: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash (or follow docs for your OS) 3. Run `hermes model` to pick Nous Portal / OpenRouter / OpenAI / your endpoint. 4. Optional: connect chat platforms + cron automations from the docs.

04

How to use it

**Simple example** You want an agent that gets better at *your* workflows over a month.

1. Install Hermes on a small VPS. 2. Talk to it from Telegram while it works on the server. 3. After a complex task, let it save a skill. 4. Next time, it reuses that skill instead of starting from zero.

05

Example prompts

  • Install Hermes and configure OpenRouter as the model provider.
  • Set a daily 9am cron: summarize my GitHub notifications to Telegram.
  • After this migration, save a reusable skill for the steps that worked.
06

Use cases and examples

  • Personal agent with long-term memory + skills
  • Cron-style reports delivered to Slack/Telegram
  • Delegating parallel subagents for research or chores
07

Prerequisites

  • A place to run it (laptop, $5 VPS, or serverless backend)
  • Model provider API key
  • Clear boundaries for what it may automate unattended
08

Tips

  • The learning loop is the point — review auto-created skills.
  • Start with chat + one cron job before wiring every messenger.