From ark
Tests the Ark Dashboard UI using Playwright, captures screenshots, and attaches them to GitHub PRs. Use when validating dashboard changes or creating PR screenshots.
How this skill is triggered — by the user, by Claude, or both
Slash command
/ark:dashboardThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Test Ark Dashboard UI with Playwright and attach screenshots to PRs.
Test Ark Dashboard UI with Playwright and attach screenshots to PRs.
Use this skill when:
CRITICAL: Before proceeding, verify Kubernetes is available:
kubectl cluster-info
If this fails, STOP and inform the user:
Cannot continue without a Kubernetes environment. Please ensure Kind or another Kubernetes cluster is running and kubectl is configured.
Ark must be deployed first. Use the ark-setup skill if needed.
Port forward the dashboard and warm up:
kubectl port-forward svc/ark-dashboard 3000:3000 -n default &
curl http://localhost:3000
Use Playwright MCP tools to navigate and screenshot:
browser_navigate - Open pagesbrowser_wait_for - Wait for elementsbrowser_click - Click elementsbrowser_take_screenshot - Capture screenshotsScreenshots save to .playwright-mcp/screenshots/. Move to ./screenshots/ for organization.
Check if user has a scratch repo:
gh repo view <USERNAME>/scratch
If not, suggest creating: scratch/pull-request-attachments/<org>_<repo>/
Upload screenshots:
cd /tmp && git clone [email protected]:<USERNAME>/scratch.git
mkdir -p scratch/pull-request-attachments/<org>_<repo>
cp ./screenshots/*.png scratch/pull-request-attachments/<org>_<repo>/
cd scratch && git add . && git commit -m "chore: screenshots for <org>/<repo> PR" && git push
Reference in PR body:

Update PR:
gh api repos/<org>/<repo>/pulls/<NUMBER> -X PATCH -f body="..."
3plugins reuse this skill
First indexed Jul 15, 2026
npx claudepluginhub mckinsey/agents-at-scale-ark --plugin arkAccumulates screenshots, videos, logs in .artifacts/<feature=branch>/ for visual regression, E2E results, and PR documentation. Generates structured reports with proof before declaring tasks complete.
Performs adversarial UI testing via browse CLI: analyzes git diffs for targeted testing or explores apps for bugs in functionality, accessibility, responsive layout, and UX heuristics.
Run and write Chainsaw e2e tests for Ark resources with mock-llm. Covers test structure, running tests, debugging failures, and patterns to avoid.