Production-grade platform engineering handbook — Kubernetes, Terraform, Flux CD, GitHub Actions, AWS, and more.
67
84%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
How to roll out platform-skills across your team's repositories — from a single repo to your entire organisation.
git installedThe fastest path. Run the installer once per repo, per tool.
git clone https://github.com/nitinjain999/platform-skills.git ~/platform-skills
cd ~/platform-skillsPick your tool:
# Claude Code — interactive plugin workflows and slash commands
claude plugin marketplace add https://github.com/nitinjain999/platform-skills
claude plugin install platform-skills
# Codex — skill invocation with $platform-skills
./install.sh --codex
# Cursor — project rules for Chat and Agent
./install.sh --cursor --target ../your-repo
# GitHub Copilot — instructions committed to the repo
./install.sh --copilot --target ../your-repo# Claude Code
claude plugin list | grep platform-skills
# Codex
codex "list skills" | grep platform-skills
# Cursor — check that rules file exists
ls your-repo/.cursor/rules/platform-skills.mdc
# Copilot — check that instructions file exists
ls your-repo/.github/copilot-instructions.md | xargs grep -l "platform-skills"Paste this into your tool of choice in the context of a real repo file:
Use $platform-skills to review the files I changed for production readiness.
Focus on ownership, blast radius, validation, rollback, and security defaults.A GitHub Actions workflow that runs install.sh across a list of repositories via a matrix strategy. One PR per target repo.
# .github/workflows/rollout-platform-skills.yml
name: Rollout platform-skills
on:
workflow_dispatch:
inputs:
tool:
description: "Tool to install (cursor | copilot | codex)"
required: true
default: cursor
repos:
description: "Comma-separated list of owner/repo"
required: true
jobs:
rollout:
runs-on: ubuntu-latest
permissions:
contents: write
pull-requests: write
strategy:
matrix:
repo: ${{ fromJson(format('["{0}"]', replace(github.event.inputs.repos, ',', '","'))) }}
fail-fast: false
steps:
- uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
with:
repository: nitinjain999/platform-skills
path: platform-skills
- uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
with:
repository: ${{ matrix.repo }}
path: target
token: ${{ secrets.ROLLOUT_TOKEN }}
- name: Install platform-skills
run: |
cd platform-skills
./install.sh --${{ github.event.inputs.tool }} --target ../target
- name: Open PR
uses: peter-evans/create-pull-request@5e914681574cf53f24b88ede31338b00d2b09b78 # v7.0.5
with:
path: target
token: ${{ secrets.ROLLOUT_TOKEN }}
commit-message: "feat: add platform-skills ${{ github.event.inputs.tool }} integration"
branch: "platform-skills/rollout-${{ github.event.inputs.tool }}"
title: "feat: add platform-skills for ${{ github.event.inputs.tool }}"
body: |
Adds platform-skills guidance for ${{ github.event.inputs.tool }}.
**What this does:** Gives ${{ github.event.inputs.tool }} access to platform engineering
patterns for Kubernetes, Terraform, Flux, GitHub Actions, AWS, and more.
**Try it:** Paste a prompt from [PROMPTS.md](https://github.com/nitinjain999/platform-skills/blob/main/PROMPTS.md)
into ${{ github.event.inputs.tool }} while reviewing any platform file.
**Rollback:** Delete `.cursor/rules/platform-skills.mdc` (or equivalent) and close this PR.Create a GitHub fine-grained Personal Access Token scoped to only the target repositories with Contents (read/write) and Pull requests (read/write) permissions. A classic PAT with repo scope is overly broad and grants unnecessary access across all your repositories.
Store it as secret ROLLOUT_TOKEN in your central tooling repo.
gh workflow run rollout-platform-skills.yml \
--field tool=cursor \
--field repos="org/repo-1,org/repo-2,org/repo-3"Each target repo gets one PR. Review and merge. To track status:
gh pr list --search "platform-skills rollout" --state openAt this scale, individual PRs are impractical. Use your existing infrastructure-as-code and repository management tooling.
Add platform-skills instructions to your organisation's default Copilot instructions. All repos in the org pick them up automatically.
platform-skills/.github/copilot-instructions.md to the organisation-level instructionsorg/platform-cursor-rulesplatform-skills/.cursor/ into it--tool cursor across all reposIf you manage repos as code, add the platform-skills install to your repo template:
# Terraform example using GitHub provider
resource "github_repository_file" "platform_skills_copilot" {
for_each = toset(var.target_repos)
repository = each.value
branch = "main"
file = ".github/copilot-instructions.md"
content = file("${path.module}/platform-skills/.github/copilot-instructions.md")
commit_message = "feat: add platform-skills copilot instructions"
}# Count repos with platform-skills installed (Copilot)
gh search code "platform-skills" --filename copilot-instructions.md \
--owner your-org --json repository | jq length
# Count repos with Cursor rules
gh search code "platform-skills" --filename "*.mdc" \
--owner your-org --json repository | jq length# Claude Code plugin — update to latest
claude plugin update platform-skills
# Cursor/Copilot — re-run install.sh from a fresh clone
git -C ~/platform-skills pull
cd ~/platform-skills && ./install.sh --cursor --target ../your-repoFor the Tier 2 matrix approach, re-trigger the workflow after each platform-skills release.
.claude-plugin
.github
commands
docs
examples
agent-self-improve
argocd
awesome-docs
aws
cloudfront
functions
lambda-edge
functions
azure
compliance
conventional-commits
datadog
llm-observability
demo
documentation
dora
dynatrace
fluxcd
github-actions
composite-actions
configure-cloud
db-migrate
docker-build-push
k8s-deploy
notify-slack
pr-comment
release-tag
security-scan
setup-env
setup-terraform
terraform-plan
helm
web-service
templates
kubernetes
kyverno
mcp
observability
openshift
pr-review
ownership
runtime-security
supply-chain
terraform
references
scripts
skills
platform-skills
tests