Install any skill in seconds. Free to start, no credit card required.
Get Started Free →CI/CD pipeline patterns for GitHub Actions, PR automation, and deployment workflows. Use when setting up CI, fixing broken pipelines, automating PR checks, or configuring deployment.
.claude/skills/kunanonj-ci-cd-patterns/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 100% | 8 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -21% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 1% | 0% |
Patterns for GitHub Actions, PR automation, and deployment workflows.
yamlname: CI on: push: branches: [main, develop] pull_request: branches: [main, develop] jobs: test: runs-on: ubuntu-latest steps: - uses: actions/checkout@v4 - name: Run tests run: ./gradlew test
yamlname: PR Check on: pull_request jobs: lint: runs-on: ubuntu-latest steps: - uses: actions/checkout@v4 - run: ./gradlew ktlintCheck test: runs-on: ubuntu-latest steps: - uses: actions/checkout@v4 - run: ./gradlew test build: runs-on: ubuntu-latest needs: [lint, test] steps: - uses: actions/checkout@v4 - run: ./gradlew build
Monitor a PR through CI, handle common failures:
gh pr checks <number>gh run rerun <run-id> --failedgh pr merge <number> --auto --squash> See workflows.md for ready-to-use GitHub Actions YAML templates.
Before deploying:
actions/cache or actions/setup-java with cache). A cold Gradle build takes 3-5 minutes, cached takes 30 seconds.needs: creates sequential dependencies. Without it, all jobs run in parallel. Use needs: [lint, test] to make build wait for checks.secrets.DB_PASSWORD and secrets.db_password are different. Match the exact name from Settings > Secrets.actions/checkout@v3 --- use v4. v3 uses Node 16 which is deprecated. v4 uses Node 20.[skip ci]) unless it's docs-only| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 20,275 | 10,733 | -47% | 1 | 1 | 0% | 3,205 | 2,781 | -13% | 0 | 0 | — |
case-06 | pass→pass | 11,674 | 9,905 | -15% | 1 | 1 | 0% | 1,977 | 2,606 | +32% | 0 | 0 | — |
case-02 | pass→pass | 10,935 | 2,253 | -79% | 1 | 1 | 0% | 1,684 | 1,235 | -27% | 0 | 0 | — |
case-03 | pass→pass | 6,857 | 3,864 | -44% | 1 | 1 | 0% | 1,697 | 1,652 | -3% | 0 | 0 | — |
case-04 | fail→pass | 9,680 | 2,906 | -70% | 1 | 1 | 0% | 1,779 | 1,398 | -21% | 0 | 0 | — |
case-05 | fail→pass | 10,640 | 5,313 | -50% | 1 | 1 | 0% | 2,050 | 1,882 | -8% | 0 | 0 | — |
case-07 | pass→pass | 8,456 | 2,832 | -67% | 1 | 1 | 0% | 1,560 | 1,301 | -17% | 0 | 0 | — |
case-08 | pass→pass | 5,517 | 2,814 | -49% | 1 | 1 | 0% | 935 | 1,161 | +24% | 0 | 0 | — |
case-09 | pass→pass | 2,873 | 1,732 | -40% | 1 | 1 | 0% | 565 | 1,128 | +100% | 0 | 0 | — |
case-10 | pass→pass | 2,202 | 1,651 | -25% | 1 | 1 | 0% | 394 | 1,045 | +165% | 0 | 0 | — |
case-11 | fail→pass | 10,165 | 4,787 | -53% | 1 | 1 | 0% | 1,840 | 1,651 | -10% | 0 | 0 | — |
case-12 | pass→pass | 10,878 | 5,139 | -53% | 1 | 1 | 0% | 1,816 | 1,634 | -10% | 0 | 0 | — |
case-13 | pass→pass | 4,676 | 3,237 | -31% | 1 | 1 | 0% | 1,024 | 1,427 | +39% | 0 | 0 | — |
case-14 | pass→pass | 10,624 | 6,211 | -42% | 1 | 1 | 0% | 1,661 | 1,872 | +13% | 0 | 0 | — |
case-15 | fail→pass | 10,175 | 9,060 | -11% | 1 | 1 | 0% | 1,699 | 1,713 | +1% | 0 | 0 | — |
case-16 | pass→pass | 7,537 | 3,230 | -57% | 1 | 1 | 0% | 1,342 | 1,324 | -1% | 0 | 0 | — |
case-17 | pass→pass | 8,340 | 1,674 | -80% | 1 | 1 | 0% | 1,383 | 1,115 | -19% | 0 | 0 | — |
case-18 | pass→pass | 14,280 | 12,468 | -13% | 1 | 1 | 0% | 2,608 | 2,899 | +11% | 0 | 0 | — |
case-19 | pass→pass | 13,539 | 4,182 | -69% | 1 | 1 | 0% | 2,377 | 1,760 | -26% | 0 | 0 | — |
case-20 | pass→pass | 11,729 | 11,035 | -6% | 1 | 1 | 0% | 2,294 | 2,963 | +29% | 0 | 0 | — |
case-21 | pass→pass | 11,626 | 11,264 | -3% | 1 | 1 | 0% | 2,279 | 3,058 | +34% | 0 | 0 | — |
case-22 | pass→pass | 7,222 | 4,366 | -40% | 1 | 1 | 0% | 1,321 | 1,631 | +23% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +23 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.