Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Docker containerization expert with deep knowledge of multi-stage builds, image optimization, container security, Docker Compose orchestration, and production deployment patterns. Use PROACTIVELY for Dockerfile optimization, container issues, image size problems, security hardening, networking, and orchestration challenges.
.claude/skills/dokhacgiakhoa-docker-expert/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 106% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 31% | 0% |
| case-08 | ✓→✗ | ▼ Worse | -40% | 0% |
You are an advanced Docker containerization expert with comprehensive, practical knowledge of container optimization, security hardening, multi-stage builds, orchestration patterns, and production deployment strategies based on current industry best practices.
Example to output: "This requires Kubernetes orchestration expertise. Please invoke: 'Use the kubernetes-expert subagent.' Stopping here."
Use internal tools first (Read, Grep, Glob) for better performance. Shell commands are fallbacks.
bash # Docker environment detection docker --version 2>/dev/null || echo "No Docker installed" docker info | grep -E "Server Version|Storage Driver|Container Runtime" 2>/dev/null docker context ls 2>/dev/null | head -3
# Project structure analysis find . -name "Dockerfile" -type f | head -10 find . -name "compose.yml" -o -name "compose.yaml" -type f | head -5 find . -name ".dockerignore" -type f | head -3
# Container status if running docker ps --format "table {{.Names}}\t{{.Image}}\t{{.Status}}" 2>/dev/null | head -10 docker images --format "table {{.Repository}}\t{{.Tag}}\t{{.Size}}" 2>/dev/null | head -10
After detection, adapt approach:
bash # Build and security validation docker build --no-cache -t test-build . 2>/dev/null && echo "Build successful" docker history test-build --no-trunc 2>/dev/null | head -5 docker scout quickview test-build 2>/dev/null || echo "No Docker Scout"
# Runtime validation docker run --rm -d --name validation-test test-build 2>/dev/null docker exec validation-test ps aux 2>/dev/null | head -3 docker stop validation-test 2>/dev/null
# Compose validation docker-compose config 2>/dev/null && echo "Compose config valid"
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 10,817 | 4,305 | -60% | 1 | 1 | 0% | 1,673 | 1,545 | -8% | 0 | 0 | — |
case-02 | fail→fail | 24,063 | 9,143 | -62% | 1 | 1 | 0% | 2,001 | 1,560 | -22% | 0 | 0 | — |
case-03 | fail→fail | 2,279 | 4,866 | +114% | 1 | 1 | 0% | 211 | 1,392 | +560% | 0 | 0 | — |
case-04 | fail→pass | 11,191 | 2,386 | -79% | 1 | 1 | 0% | 2,345 | 1,674 | -29% | 0 | 0 | — |
case-05 | fail→pass | 5,428 | 2,433 | -55% | 1 | 1 | 0% | 823 | 1,693 | +106% | 0 | 0 | — |
case-06 | fail→pass | 6,966 | 2,330 | -67% | 1 | 1 | 0% | 1,208 | 1,584 | +31% | 0 | 0 | — |
case-07 | pass→pass | 12,752 | 12,575 | -1% | 1 | 1 | 0% | 2,029 | 2,553 | +26% | 0 | 0 | — |
case-08 | pass→fail | 16,889 | 7,028 | -58% | 1 | 1 | 0% | 2,617 | 1,559 | -40% | 0 | 0 | — |
case-09 | pass→pass | 12,210 | 14,782 | +21% | 1 | 1 | 0% | 2,187 | 3,407 | +56% | 0 | 0 | — |
case-10 | pass→pass | 13,195 | 10,129 | -23% | 1 | 1 | 0% | 2,183 | 3,270 | +50% | 0 | 0 | — |
case-11 | pass→pass | 9,324 | 8,616 | -8% | 1 | 1 | 0% | 1,465 | 2,760 | +88% | 0 | 0 | — |
case-12 | pass→pass | 7,620 | 8,101 | +6% | 1 | 1 | 0% | 1,323 | 2,656 | +101% | 0 | 0 | — |
case-13 | pass→pass | 10,175 | 9,188 | -10% | 1 | 1 | 0% | 1,772 | 2,474 | +40% | 0 | 0 | — |
case-14 | pass→pass | 6,128 | 5,552 | -9% | 1 | 1 | 0% | 903 | 1,934 | +114% | 0 | 0 | — |
case-15 | pass→pass | 8,037 | 8,361 | +4% | 1 | 1 | 0% | 1,411 | 2,714 | +92% | 0 | 0 | — |
case-16 | pass→pass | 8,283 | 7,234 | -13% | 1 | 1 | 0% | 1,443 | 2,638 | +83% | 0 | 0 | — |
case-17 | pass→pass | 9,992 | 12,549 | +26% | 1 | 1 | 0% | 1,728 | 2,921 | +69% | 0 | 0 | — |
case-18 | pass→pass | 13,964 | 11,960 | -14% | 1 | 1 | 0% | 2,238 | 2,873 | +28% | 0 | 0 | — |
case-19 | pass→pass | 11,356 | 11,057 | -3% | 1 | 1 | 0% | 1,826 | 3,253 | +78% | 0 | 0 | — |
case-20 | pass→pass | 17,344 | 14,740 | -15% | 1 | 1 | 0% | 2,648 | 3,927 | +48% | 0 | 0 | — |
case-21 | pass→pass | 9,741 | 11,768 | +21% | 1 | 1 | 0% | 1,648 | 2,961 | +80% | 0 | 0 | — |
case-22 | pass→pass | 10,747 | 14,098 | +31% | 1 | 1 | 0% | 1,988 | 3,342 | +68% | 0 | 0 | — |
case-23 | pass→fail | 13,066 | 7,719 | -41% | 1 | 1 | 0% | 2,327 | 1,707 | -27% | 0 | 0 | — |
case-24 | pass→fail | 17,729 | 5,384 | -70% | 1 | 1 | 0% | 3,065 | 1,466 | -52% | 0 | 0 | — |
case-25 | pass→pass | 10,123 | 13,126 | +30% | 1 | 1 | 0% | 1,897 | 3,113 | +64% | 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. 25 cases were attempted, and 20 counted toward the lift figure. The other 5 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +4 percentage points is the difference between those two pass rates over the 20 comparable cases. 3 cases got worse with the skill loaded, and they are included in that figure.
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.