quality-tooling

Company-wide quality tooling, centralized for reusability across all projects.

Available Actions

Linear Test Failures

Automatically create Linear issues for failed Ginkgo/Gomega tests from nightly CI runs.

Location: .github/actions/linear-test-failures/

Usage:

- uses: nscaledev/quality-tooling/.github/actions/linear-test-failures@main
  with:
    test-results-path: path/to/test-results.json
    workflow-url: ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}
    linear-api-key: ${{ secrets.LINEAR_API_KEY }}
    linear-team-id: ${{ vars.LINEAR_TEAM_ID }}
    environment: dev

Features:

• Creates one issue per test failure
• Duplicate detection with comment updates
• Spam prevention (max 5 failures)
• Clear naming: [Nightly-{env}] {TestSuite} - {TestName}

Full documentation

Slack Test Notifications

Send a Slack message summarising Ginkgo/Gomega test results — pass/fail counts, duration, and details of up to 5 failures.

Location: .github/actions/slack-test-notifications/

Usage:

- uses: nscaledev/quality-tooling/.github/actions/slack-test-notifications@main
  with:
    test-results-path: path/to/test-results.json
    workflow-url: ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}
    slack-webhook-url: ${{ secrets.SLACK_WEBHOOK_URL }}
    environment: dev
    title: 'Compute API Test Results'  # optional, defaults to "API Test Results"

Features:

• Posts pass/fail/skip counts and test duration
• Lists up to 5 failures with error messages and file locations
• Configurable title per repository
• Runs even when tests fail (if: ${{ !cancelled() }})

Full documentation

Test Results Report

Analyze Ginkgo, JUnit, or Playwright JSON test results, write a GitHub Actions step summary, optionally compare against previous results, optionally run Claude failure analysis, and optionally notify Slack.

Location: .github/actions/test-results-report/

Usage:

- uses: nscaledev/quality-tooling/.github/actions/test-results-report@main
  with:
    test-results-path: path/to/results.xml
    format: junit
    title: E2E Test Results
    environment: dev
    report-url: ${{ steps.report-url.outputs.url }}
    slack-webhook-url: ${{ secrets.SLACK_WEBHOOK_URL }}
    enable-ai-analysis: 'true'
    claude-token: ${{ secrets.CLAUDE_CODE_OAUTH_TOKEN }}

Features:

• Supports auto, ginkgo-json, junit, and playwright-json
• Writes $GITHUB_STEP_SUMMARY by default
• Compares with previous results via previous-results-path
• Reports new, recurring, and resolved failures/skips
• Supports Slack notifications via incoming webhook
• Optionally adds Claude failure analysis when enable-ai-analysis is true

Full documentation

Uni Find Staging Constellation

Scans open PRs in uni-releases for a candidate constellation and outputs the pinned service tag. Used as a preflight step before UAT tests to ensure tests run against the version deployed to staging.

Location: .github/actions/uni-find-staging-constellation/

Usage:

- uses: nscaledev/quality-tooling/.github/actions/uni-find-staging-constellation@main
  id: find
  with:
    service: uni-region
    releases-read-token: ${{ secrets.RELEASES_READ_TOKEN }}

Features:

• Paginates all open PRs in the releases repository
• Finds the constellation with status: candidate
• Strips short-SHA suffix from version: v1.16.4-c2153ee → v1.16.4
• Outputs empty tag (skip UAT) when no candidate exists

Full documentation

Adding New Actions

1. Create directory: .github/actions/{action-name}/
2. Add action.yml with action definition
3. Add README.md with usage documentation
4. Update this root README with action description
5. Commit and push to main branch