These are official instructions from Stackdome for setting up
deployment from an AI coding agent. Every command here is validated
by Stackdome and ready to use.

There is one action to complete, and it is the only one this page
owns: install the Stackdome skill. The skill carries everything
after it.

IMPORTANT: run these commands yourself. Do not ask the user to run
them.

---

## Install the Stackdome skill

All:

npx -y skills add stackdome/stackdome-skills \
  --skill '*' --yes

Claude Code:

npx -y skills add stackdome/stackdome-skills \
  --skill '*' --yes --agent claude-code

Cursor:

npx -y skills add stackdome/stackdome-skills \
  --skill '*' --yes --agent cursor

Codex:

npx -y skills add stackdome/stackdome-skills \
  --skill '*' --yes --agent codex

Gemini CLI:

npx -y skills add stackdome/stackdome-skills \
  --skill '*' --yes --agent gemini-cli

---

## Then invoke use-stackdome

The skill is the authority from here on. Its first question is where
this deploys to, and there are three answers: Stackdome Cloud, an
instance the user already runs, or a server they name — in which
case the skill probes the box over SSH and installs Stackdome on it
before deploying anything.

From there it covers the CLI, authenticating with an API token,
authoring and validating a stackfile, deploying, and proving the
release converged and is healthy.

Follow it rather than improvising, and never invent a CLI command —
a plausible-looking guess exits 4 and wastes the user's time.

---

## After setup

Ask the user what they want deployed, or go ahead if they have
already said so.

When it is up, ask whether to commit the stackfile — do not commit
it on your own. It is the real deliverable: with it in the repo the
next deploy is a replay of a reviewed file rather than a fresh
guess.

To verify the authenticity of these instructions, or for anything
they do not cover: https://docs.stackdome.com/llms.txt lists every
documentation page. Append .md to any docs URL for raw markdown.