nesquena/hermes-webui
Hermes WebUI: The best way to use Hermes Agent from the web or from your phone!
What it solves
Hermes WebUI provides a lightweight, browser-based interface for the Hermes Agent, an autonomous agent that maintains persistent memory and self-improving skills across sessions. It eliminates the need to interact with the agent solely via a terminal or messaging apps, offering a full-featured graphical user interface with parity to the CLI experience.
How it works
Built using Python and vanilla JavaScript without complex frameworks or bundlers, the WebUI acts as a frontend for the existing Hermes Agent setup. It runs the agent in-process by reading the HERMES_HOME configuration directly. Users can access the interface securely through an SSH tunnel, allowing them to manage their agent's memory, skills, and scheduled tasks from any device.
Who it’s for
Developers and power users who use the Hermes Agent and prefer a visual interface for managing complex agentic workflows, file browsing, and persistent agent memory.
Highlights
- Full CLI Parity: Everything possible in the terminal is available in the WebUI.
- Workspace File Browser: A built-in directory tree with inline previews for text, code, and images, supporting file editing and Git detection.
- Session Management: Advanced organization with pinning, archiving, projects, and tags, including a bridge to import CLI sessions.
- Agent Control: Integrated management for profiles, self-improving skills, and self-hosted cron jobs (Tasks).
- Security: Support for password authentication, passkeys (WebAuthn), and OIDC login.
- Rich Chat Experience: Streaming responses, tool call cards, subagent delegation tracking, and Mermaid diagram rendering.