GlueOps Deploy

This skill is a stub. The real instructions live in a single canonical LLM.md file hosted on GitHub. Fetch it, then follow it.

What to do

1. Fetch the canonical guide:

   https://raw.githubusercontent.com/GlueOps/glueops-dev/main/LLM.md

   Use whatever fetch tool your harness exposes (e.g. WebFetch, curl via a shell tool, an HTTP tool). If fetching fails, tell the user and ask them to paste the contents of that URL.

2. Read the fetched document in full before generating any files or commands.

3. Follow it. The document specifies:

   • The two repos involved (deployment-configurations + the app source repo)
   • The directory layout under apps/<app-name>/
   • The shape of base-values.yaml and envs/<env>/values.yaml
   • When to use Standard Ingress vs Traefik IngressRoute CRDs
   • Helm template expressions to use instead of hardcoding
   • Secrets via externalSecret + OpenBao
   • Verification (no kubectl — use curl and the ArgoCD dashboard)
   • Common gotchas and the questions to ask the user before generating files

4. Do not generate raw Kubernetes manifests or kubectl commands. Everything goes through values files in the user's deployment-configurations repo.

Notes

• The canonical guide is updated in-place on main. Always fetch fresh — do not rely on a cached copy from a previous session.
• This stub is harness-agnostic in spirit: any agent (Claude Code, Cursor, Continue, an OpenAI Assistant, a custom script) can be pointed at the same URL and get the same instructions. The name / description frontmatter above follows the Claude Code skill format; other harnesses may need their own equivalent wrapper around the same fetch step.