agentleFS
jbouderjbouder/skills/new-nebari-app/SKILL.md

new-nebari-app

Skill1 starsChanged 25 days agoMITView on GitHub

Scaffold a static web app (plain HTML/CSS/JS, no build step) in the layout the Nebari Apps Pack expects, with a nebari-app.yaml launch manifest, and launch it on the cluster through the nebari-apps MCP server. Use when the user wants a static page or prototype deployed to a Nebari cluster ("scaffold a static Nebari app", "make a page and deploy it to Nebari"), or says "launch it" about an app directory containing nebari-app.yaml. Not for React/FastAPI codebases — use new-project for those.

The file

We split this skill into the rules it gives an agent

5 rules. About 0 are common, 0 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Every app is a directory with a nebari-app.yaml next to the content:

    line 15
  2. Very specific:

    When the user asks to launch an app directory, read its nebari-app.yaml and drive the nebari-apps MCP tools:

    line 46
  3. Very specific:

    If a tool returns an auth error, call authenticate and follow its instructions (show the user the verification URL + code, then retry).

    line 50
  4. Very specific:

    source.type: files: read every text file under files.path (relative to the manifest) and pass them as inline_files ({relative/path: …

    line 54
  5. Very specific:

    Keep nebari-app.yaml in version control — it is also the GitOps artifact: piped through App metadata, kubectl apply works on the rendered …

    line 69

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Scaffold and launch Nebari apps

Apps on a Nebari cluster are launched from one declarative contract. This skill scaffolds static app directories in that contract's layout — content plus a nebari-app.yaml manifest — and maps the manifest onto the nebari-apps MCP server's launch_app tool so "generate, then launch it" is one smooth flow.

The layout

Every app is a directory with a nebari-app.yaml next to the content:

docs-site/
  nebari-app.yaml
  index.html
  styles.css

The manifest maps 1:1 onto the pack's App resource — see references/manifest.md for the full schema.

Scaffolding

  1. Pick names. Directory and app name: lowercase letters, digits, hyphens (max 53 chars). Default the subdomain to the app name — the app's URL becomes http(s)://<subdomain>.<appsDomain>.
  2. Scaffold from the template in assets/static/: a real index.html (+ styles.css), manifest with source.type: files. Author actual files; never hand-write inline YAML content.
  3. Write nebari-app.yaml with real values, not placeholders. Ask about access (public vs group-restricted) if the user hasn't said.
  4. Tell the user how to launch, verbatim:

    Say "launch it" and I'll deploy it to the cluster via the nebari-apps MCP server (https://apps.<cluster-domain>/mcp).

Launching ("launch it")

When the user asks to launch an app directory, read its nebari-app.yaml and drive the nebari-apps MCP tools:

  1. describe_cluster — confirm the target namespace is available and get appsDomain. If a tool returns an auth error, call authenticate and follow its instructions (show the user the verification URL + code, then retry).
  2. Map the manifest to a launch_app call — the full mapping table is in references/manifest.md. The common case:
    • source.type: files: read every text file under files.path (relative to the manifest) and pass them as inline_files ({relative/path: content}). Must include index.html; text assets only; keep the total under ~900KB — if larger, push the directory to git and use a git source instead.
  3. launch_app is idempotent on (namespace, name) — re-launching an existing app updates it, so retrying or iterating is safe.
  4. Poll get_app_status until phase is Running (typically seconds). If it lands in Failed, read get_app_logs and the status message, fix, re-launch.
  5. Report the app's url back to the user.

Conventions (do not deviate)

  • The platform serves content with nginx on port 8080. The gateway terminates TLS and, for non-public apps, enforces Keycloak SSO — never add auth code to the app itself.
  • Keep nebari-app.yaml in version control — it is also the GitOps artifact: piped through App metadata, kubectl apply works on the rendered form.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.