Hermes HUD
A local browser dashboard for the Hermes AI agent that combines live chat, memory inspection, model cost analytics and redacted session exports.
Hermes HUD is a browser dashboard for people who already run the Hermes AI agent and want to inspect its activity and persistent memory. It requires a running agent and reads its local data.
- Agent overview: Follow health, spending, model usage and provider problems through live updates, with views for memory, skills, sessions, scheduled jobs and projects.
- Chat and diagnostics: Talk to the agent in the dashboard, investigate health issues, and compare token use and costs by model.
- Tools and plugins: See whether web search, image generation, speech and browser automation use Nous Tool Gateway or direct keys, and enable, disable or update installed extensions.
- Session replays: Inspect run timelines and export JSON, Markdown, standalone HTML or PNG share cards; default Safe Share Mode redacts tool arguments, terminal output, reasoning and sensitive fields.
The dashboard is open source under the MIT license and runs locally with Python and Node.js. It accepts connections from the same machine by default. It works independently of the hermes-hud terminal companion, so you don't need that interface to use the browser dashboard.
Replay exports stay on your machine by default. Optional remote publishing sends sanitized replay artifacts to a Git repository for a static gallery, including GitHub Pages, only when you trigger a sync.