Skip to content

Hermes WebUI Agent - Setup Guide

Source: reason-machines/hermes-skills (240+ installs) Category: Hermes Agent / UI & Deployment Quality Tier: 🟡 Beta

Expert skill for deploying, configuring, and using Hermes WebUI - a lightweight web interface for Hermes Agent built on Flask. Provides browser-based access with three-panel layout (sessions sidebar, chat center, workspace file browser), persistent memory, scheduled cron jobs, and 10+ messaging platform integrations.


Installation

npx skills add reason-machines/hermes-skills --skill hermes-webui-agent
git clone https://github.com/nesquena/hermes-webui.git
cd hermes-webui
python3 bootstrap.py
# Auto-detects Hermes Agent, creates venv, starts on http://localhost:8787

Docker

git clone https://github.com/nesquena/hermes-webui
cd hermes-webui
cp .env.docker.example .env
docker compose up -d
# Open http://localhost:8787

Prerequisites

Requirement Details
Python 3.10+ for native install
Docker For containerized deployment
Hermes Agent Installed or auto-detected by bootstrap
Model Provider OpenAI, Anthropic, DeepSeek, OpenRouter, or local Ollama

Key Capabilities

Browser-Based Agent Access

Full three-panel WebUI with sessions sidebar, chat center with markdown + syntax highlighting, and workspace file browser. Supports real-time SSE streaming, tool call rendering, and multi-session management.

Scheduled Jobs Dashboard

Create, view, and manage cron-style jobs from the WebUI Control Center. Jobs run offline and deliver results via Telegram, Discord, Slack, Signal, or email.

Messaging Platform Integration

Configure 10+ messaging platforms directly from WebUI. Supports Telegram bots, Discord webhooks, Slack webhooks, Signal (via signal-cli), and email delivery for scheduled job output.

Self-Hosted with Security

Password protection, SSH tunneling for remote access, and reverse proxy support (Nginx example included). All conversations and memory stay on your hardware.

Daemon Lifecycle Management

Full ctl.sh daemon control: start, stop, restart, status, and log viewing. Writes logs to ~/.hermes/webui.log and PID to ~/.hermes/webui.pid.


Quick Start

# Clone and bootstrap (one-command setup)
git clone https://github.com/nesquena/hermes-webui.git && cd hermes-webui
python3 bootstrap.py

# Or Docker
docker compose up -d

# Access at http://localhost:8787

Remote Access via SSH Tunnel

# On your local machine
ssh -L 8787:localhost:8787 user@your-server.com
# Then open http://localhost:8787

Password Protection (Required for Network Exposure)

echo "HERMES_WEBUI_PASSWORD=your-strong-password" >> .env
docker compose up -d --force-recreate

Verification

# Check health endpoint
curl http://localhost:8787/health

# Check daemon status (if using ctl.sh)
./ctl.sh status

# Check logs
./ctl.sh logs --lines 50

Notes

  • Bootstrap auto-detects Hermes Agent installation - no manual path configuration needed
  • Docker supports single-container, two-container (agent+webui isolated), and three-container (agent+dashboard+webui) setups
  • Supports Podman with some limitations (known issue with shared .hermes in Podman 3.4 - upgrade to Podman 4+)
  • Workspace file browser mirrors HERMES_WEBUI_DEFAULT_WORKSPACE directory
  • For production, use reverse proxy (Nginx) with HTTPS and password protection enabled

Integration with CorpusIQ

CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.

This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.

See what CorpusIQ connects

AI answers on your data. See where every number comes from.

Connect your tools and ask in ChatGPT, Claude, or Perplexity. Enterprise-grade answers in minutes. From $29.95/month.

Start free, no credit card
Contact Terms Privacy Pricing