Track 1 — Hermes (web dashboard, in Docker)

Goal: open the Hermes dashboard in your browser, point it at Ollama Cloud, and tour the files that make Hermes tick — Skills, Profiles, Memory, Tools.

This track runs as a container in the kit (no host install). ./bootstrap.sh already started it.

Hermes is pre-1.0 and moves fast — if a command here differs from the live docs, the docs win. Steps marked # VERIFY should be checked against the site. Docs: https://hermes-agent.nousresearch.com/docs/user-guide/features/web-dashboard

1. Open the dashboard

Dashboard: http://localhost:9119
Log in: username trainee · password hermes

That’s the official nousresearch/hermes-agent image running the gateway + web dashboard. On first start it populates a real home into ./hermes-home/ on the host: a 16 KB config.yaml, a SOUL.md, bundled skills/ (≈17 categories), plus memories/, sessions/, cron/, kanban/, … — real files you can read and edit while it runs.

See the files in VS Code: they’re also mounted into the browser VS Code at http://localhost:8080 — the Explorer shows hermes-home alongside latex-starter, so you can browse and edit SOUL.md/skills//memories/ there.

If 9119 doesn’t load yet, give the container a few seconds on first build, or check it’s up: docker compose ps / docker compose logs hermes.

2. It’s already on Ollama Cloud

./bootstrap.sh configures Hermes to use Ollama Cloud with your OLLAMA_API_KEY (from .env) — the default model is glm-5.1:cloud and the helper models are set too, so the chat just works.

Start a new chat (the New chat button / model selector, top-right) — an old session may still show a different model.
To change the model: the dashboard’s model selector, or Config. Browse tool-capable cloud models at https://ollama.com/search?c=cloud.

Cloud models handle their own context window — no local-model / 64K-context setup needed.

3. Guided tour (2–3 minutes each)

Open these screens in the dashboard and connect each to the lesson site, then find the matching file on the host in ./hermes-home/:

Dashboard screen	What you’re looking at	On disk (`./hermes-home/`)
Skills	reusable `SKILL.md` procedures + toolsets	`skills/`
Profiles	switchable, isolated configs	(per-profile config)
Memory	persistent memory + provider choice	`memories/`
Tools	the actions the agent can take	—
Chat	talk to the agent (embedded terminal)	`sessions/`

Also open ./hermes-home/SOUL.md in your editor — change a line, start a new chat in the dashboard, and watch the behaviour shift. And note AGENTS.md: project-scoped instructions that complement the global SOUL.md.

Done when: the dashboard is open, talking to Ollama Cloud, and you can point to where Skills / Profiles / Memory live on disk and say what each is.

Your Track 1 build — automate the job

The tour above is the warm-up. Track 1’s actual goal is to automate a job: get Hermes to do what the LaTeX web UI does (topic → image), but unattended, on a daily schedule — using a skill + a cron job + memory. That’s its own lesson: Automate with Hermes. (Track 3 builds the latex-mcp infrastructure; Track 1 automates the basic job.)

Alternative — native Hermes Desktop

Prefer the native desktop app (streaming previews, voice, no browser)? Install it from https://hermes-agent.nousresearch.com/docs/getting-started/installation (or curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash, then hermes desktop). The concepts are identical; only the front-end differs.

Concepts behind this: What is Hermes, Skills, Profiles & SOUL.md, AGENTS.md, Memory.