Summary

Hermes HUD UI provides a browser-based dashboard for monitoring the internal state and operations of the Hermes AI agent. It visualizes everything the agent knows about itself, including its identity, conversations, memories, skills, active projects, tool usage patterns, and token costs. This tool is designed for users of the Hermes AI agent who require a comprehensive, real-time overview of their agent's "consciousness" and operational status, offering a richer, browser-accessible experience compared to its TUI counterpart.

How It Works

The system employs a dual-architecture approach: a FastAPI backend written in Python and a React frontend utilizing Vite and SWR. The backend reads agent data directly from the ~/.hermes/ directory through dedicated collectors. It features intelligent caching with mtime-based invalidation for sessions, skills, and patterns, coupled with a file watcher to detect changes. Real-time updates are pushed to the frontend via WebSockets, ensuring the dashboard reflects agent activity instantly. The frontend uses SWR for efficient data fetching and background updates, presenting stale data during refreshes to avoid UI blinking.

Quick Start & Requirements

• Primary install: Clone the repository, navigate into the directory, create and activate a Python 3.11 virtual environment, and run ./install.sh. The application is then started with hermes-hudui. Access it at http://localhost:3001.
• Prerequisites: Python 3.11+, Node.js 18+ (for frontend build), and a running Hermes agent with data located in ~/.hermes/.
• Links: http://localhost:3001

Highlighted Details

• Real-time data updates via WebSocket connection with automatic reconnection and exponential backoff.
• Smart caching mechanism with automatic invalidation based on file modification times.
• Four distinct visual themes (Neural Awakening, Blade Runner, fsociety, Anime) with an optional CRT scanline overlay, switchable via keyboard shortcuts.
• Comprehensive visualization of agent's state, including memory capacity, user profile, API key status, service health, learning progress, project status, tool usage, and daily activity sparklines.
• Detailed token cost tracking per model, with support for various providers (Anthropic, OpenAI, DeepSeek, xAI, Google) and fallback pricing for unsupported models.
• Offers features beyond the TUI version, such as dedicated Memory, Skills, and Sessions tabs, and a command palette.

Maintenance & Community

No specific details regarding maintainers, community channels (e.g., Discord, Slack), or roadmaps were found in the provided README.

Licensing & Compatibility

• License: MIT.
• Compatibility: The MIT license is permissive and generally allows for commercial use and integration into closed-source projects without significant restrictions.

Limitations & Caveats

The application requires a pre-existing, running Hermes agent and direct access to its data directory. Native Windows support is not provided; users must utilize WSL (Windows Subsystem for Linux). Frontend development and building necessitate Node.js. Token cost calculations rely on hardcoded pricing, which may require manual updates for new models or pricing changes.