Deploy and Host Hermes Agent (Full) on Railway

Hermes is a self-improving AI agent by Nous Research that connects to your messaging channels, learns from every interaction, creates its own skills, and gets more capable over time. This Full template runs the agent plus the first-party Nous hermes dashboard — built from source — so you can configure and monitor everything from your browser.

About Hosting Hermes Agent (Full)

The Full variant serves the official Hermes web dashboard behind basic auth (not a third-party UI): configure LLM providers, channels, and tools; watch live gateway logs; and start/stop the agent — all point-and-click. Alongside it runs the same always-on hermes gateway that answers Telegram, Discord, and Slack 24/7. Config, skills, memories, and sessions persist on a Railway volume (/opt/data) so they survive redeploys. Steady memory is ~300 MB (dashboard + gateway) with no headless browser bundled — lean for what it offers.

Why Deploy Hermes Agent (Full) on Railway?

• Official dashboard, from source — the genuine Nous hermes dashboard, not a community wrapper.
• Point-and-click setup — pick providers/channels/tools and watch logs in the browser; no terminal required.
• Self-improving — agent-curated memory, autonomous skill creation, cron jobs.
• Lean for a UI build — ~300 MB steady, no Chromium/Playwright.
• Your keys, your data — nothing is bundled and no credentials ship with the template.

Want the lowest possible RAM and don't need a web UI? Deploy the Slim variant (~112 MB, CLI + gateway) — also on this marketplace.

Common Use Cases

• A browser-managed personal assistant you configure without touching a terminal.
• A team agent where non-CLI users tweak providers/channels via the dashboard.
• A self-improving bot with live observability of its gateway and logs.

Dependencies for Hermes Agent (Full) Hosting

On deploy you choose your dashboard login — just a HERMES_DASHBOARD_BASIC_AUTH_USERNAME and HERMES_DASHBOARD_BASIC_AUTH_PASSWORD. The session-signing secret is generated for you on first boot and persisted to the volume, so there's no opaque value to fill in. Then give it an LLM: set an API-key variable (OPENAI_API_KEY, OPENROUTER_API_KEY, …) or sign in with a ChatGPT / Codex subscription — or just configure a provider from the dashboard after it boots.

Deployment Dependencies

• Hermes Agent — open source by Nous Research; the dashboard is built from upstream source.