Hermes WebUI Brings Browser and Mobile Access to Hermes Agent

Hermes WebUI is a lightweight, dark-themed browser interface designed to interact with the Hermes Agent. It aims to provide parity with the command-line interface (CLI) experience using Python and vanilla JavaScript, without requiring complex build steps or heavy frameworks.

Core Architecture and Layout

The interface features a three-panel layout:

• Left Sidebar: Manages sessions, navigation, and access to the Hermes Control Center.
• Center Area: Dedicated to the chat interface.
• Right Sidebar: Provides workspace file browsing capabilities.

Controls for models, profiles, and workspaces remain visible in the composer footer during message composition, accompanied by a circular context ring indicating token usage.

Key Features and Capabilities

• Persistent Memory: Retains context across sessions and supports a skills system that saves reusable procedures.
• Self-Hosted Scheduling: Runs cron jobs offline and delivers notifications to platforms such as Telegram, Discord, Slack, and email.
• Multi-Provider Support: Compatible with OpenAI, Anthropic, Google, DeepSeek, OpenRouter, and custom endpoints.
• Workspace Management: Offers directory trees, inline previews, file editing, and Git branch detection.
• Voice Integration: Includes microphone support via the Web Speech API for live interim transcription.

Setup and Verification

Users can run the setup locally using provided scripts or Docker containers. To verify installation, check local server logs or use the health check endpoint as documented in the official repository. Native Windows is not directly supported by the primary bootstrap script, though community guides exist for alternative configurations.