Hermiss: automation for Hermes Agent

Q: How do I install Hermiss?

Clone the repository, copy .env.example to .env and run docker compose up -d, then open the panel at http://127.0.0.1:8788. On Windows you can double-click the one-click deploy batch file instead.

Self-hosted virtual companion built on Hermes Agent with WeChat access, memory and personas

What Hermiss does

Hermiss is a self-deployed companion assistant built on Hermes Agent. It chats with you through WeChat, bound by scanning a QR code in a local web panel, and supports long-term memory of preferences, state and recent events, persona management and a sticker system. Personas are edited through SOUL.md and USER.md, and an AI can draft a persona for you.

This is the single-user edition: you log in to the panel for your own container, with no multi-user admin side. The model defaults to DeepSeek and can be changed in the panel by provider, model, base URL and API key. After a conversation ends, it can send a proactive follow-up based on recent context. Deployment uses Docker Compose, with a double-click batch file on Windows, and the stack consists of a Hermiss image, a panel image and a Milvus vector database. Updates keep personas, memory, configuration and stickers in Docker volumes. The README is in Chinese.

Key features

WeChat binding by QR code from a local web panel
Configurable model provider, model, base URL and API key, with DeepSeek as default
Long-term memory for preferences, state and recent events
SOUL.md and USER.md persona editing, with AI-drafted personas
Sticker system with categories, upload, preview and usage records
Proactive follow-up messages after a conversation

When to use it

Run a private companion chatbot on your own machine that talks over WeChat
Tune a persona and memory for a single user
Deploy and update the whole stack on Windows with a double-click script

Who it is for: Chinese-speaking users who want a self-hosted companion chatbot on WeChat built on Hermes Agent.

How it fits with Hermes Agent

It is built on Hermes Agent and keeps its SOUL.md and USER.md persona files, packaged with a panel and Docker images.

How to install Hermiss

These commands are copied from the project's README. Check the repository for the latest steps before you run them.

git clone https://github.com/LinMuMuPro/hermiss.git
cd hermiss
cp .env.example .env
docker compose up -d

Requirements: Docker (Docker Desktop on Windows) and a model API key; the panel runs at http://127.0.0.1:8788 by default

Note: It is a single-user edition, and the repository has no license file.

FAQ

What is Hermiss?

Hermiss is a self-hosted virtual companion assistant built on Hermes Agent. It talks to you through WeChat and supports long-term memory, personas, stickers and proactive replies.

How do I install Hermiss?

Clone the repository, copy .env.example to .env and run docker compose up -d, then open the panel at http://127.0.0.1:8788. On Windows you can double-click the one-click deploy batch file instead.

What do I need to run Hermiss?

You need Docker and an API key for a model provider, with DeepSeek as the default. You also bind a WeChat account by scanning a QR code in the panel.