Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Workflows für Shell-Scripting, Automatisierung, Parsing, CI-Tooling.
.claude/skills/shadd0wtaka-bash-automation-scripting-skill/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 64% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -3% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 22% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 26% | 0% |
| case-17 | ✗→✓ | ▲ Improved | 39% | 0% |
Workflows für Shell-Scripting, Automatisierung, Parsing, CI-Tooling.
bash#!/usr/bin/env bash set -euo pipefail IFS=$'\n\t' # Config VERSION="1.0.0" LOG_FILE="/tmp/script.log" # Colors RED='\033[0;31m'; GREEN='\033[0;32m'; YELLOW='\033[1;33m'; NC='\033[0m' info() { echo -e "${GREEN}[✓]${NC} $*"; } warn() { echo -e "${YELLOW}[!]${NC} $*"; } error() { echo -e "${RED}[✗]${NC} $*" >&2; exit 1; } # Usage usage() { cat <<EOF Usage: $(basename "$0") [options] Options: -h, --help Show help -v, --verbose Verbose output -o, --output Output file EOF exit 0 } [[ $# -eq 0 ]] && usage while [[ $# -gt 0 ]]; do case $1 in -h|--help) usage ;; -v|--verbose) VERBOSE=1 ;; -o|--output) OUTPUT="$2"; shift ;; *) error "Unknown: $1" ;; esac shift done info "Starting..." trap 'warn "Interrupted"; exit 1' INT TERM trap 'info "Cleanup done"' EXIT
bash# Run with timeout timeout 30 command || { error "Command timed out"; } # Retry loop for i in {1..3}; do command && break || sleep $((i * 2)) done # Log everything exec 1> >(tee -a "$LOG_FILE") 2>&1
bash# API Response parsen curl -s https://api.example.com/data | jq -r ' .items[] | select(.status == "active") | {name, id, created} ' # CSV → JSON jq -R 'split(",") | {name: .[0], age: .[1] | tonumber}' < data.csv | jq -s
bash# Config lesen yq '.services.web.image' docker-compose.yml # Value setzen yq -i '.version = "2.0"' config.yaml
bash# Bulk cleanup docker ps -aq --filter "status=exited" | xargs -r docker rm docker images -q --filter "dangling=true" | xargs -r docker rmi # Bulk exec for c in $(docker ps -q --filter "name=web"); do docker exec "$c" kill -USR2 1 # reload done # Healthcheck loop until curl -sf http://localhost:8080/health; do echo "Waiting..."; sleep 2 done info "Service ready"
bash# Bulk branch cleanup git branch --merged main | grep -v "main\|*" | xargs -r git branch -d # Auto-commit git add -A && git commit -m "chore: auto-update $(date +%Y-%m-%d)" || true # Status for all repos for d in ~/projects/*/; do (cd "$d" && echo "=== $d ===" && git status -s) done
bash# Semantic Version from Git git_describe_version() { local tag=$(git describe --tags --abbrev=0 2>/dev/null || echo "0.0.0") local commit=$(git rev-list --count HEAD 2>/dev/null || echo "0") echo "${tag}-build.${commit}" } # Check if CI is_ci() { [ -n "${CI:-}" ] || [ -n "${GITHUB_ACTIONS:-}" ] || [ -n "${GITLAB_CI:-}" ] } # Parallel execution run_parallel() { local pids=() for cmd in "$@"; do (eval "$cmd") & pids+=($!) done for pid in "${pids[@]}"; do wait "$pid"; done }
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 18,996 | 17,130 | -10% | 1 | 1 | 0% | 3,535 | 4,537 | +28% | 0 | 0 | — |
case-02 | fail→pass | 8,351 | 7,651 | -8% | 1 | 1 | 0% | 1,538 | 2,526 | +64% | 0 | 0 | — |
case-03 | fail→pass | 20,414 | 16,149 | -21% | 1 | 1 | 0% | 4,351 | 4,229 | -3% | 0 | 0 | — |
case-04 | pass→pass | 6,946 | 6,314 | -9% | 1 | 1 | 0% | 1,385 | 2,185 | +58% | 0 | 0 | — |
case-05 | pass→pass | 11,734 | 8,843 | -25% | 1 | 1 | 0% | 2,106 | 2,666 | +27% | 0 | 0 | — |
case-06 | pass→pass | 11,779 | 9,900 | -16% | 1 | 1 | 0% | 2,281 | 2,835 | +24% | 0 | 0 | — |
case-07 | fail→pass | 7,595 | 3,884 | -49% | 1 | 1 | 0% | 1,406 | 1,715 | +22% | 0 | 0 | — |
case-08 | pass→pass | 9,418 | 4,721 | -50% | 1 | 1 | 0% | 1,682 | 1,926 | +15% | 0 | 0 | — |
case-09 | fail→fail | 11,068 | 8,857 | -20% | 1 | 1 | 0% | 2,049 | 2,613 | +28% | 0 | 0 | — |
case-10 | fail→pass | 11,401 | 8,511 | -25% | 1 | 1 | 0% | 2,054 | 2,580 | +26% | 0 | 0 | — |
case-11 | pass→pass | 11,661 | 8,026 | -31% | 1 | 1 | 0% | 2,254 | 2,515 | +12% | 0 | 0 | — |
case-12 | pass→pass | 7,146 | 6,261 | -12% | 1 | 1 | 0% | 1,292 | 2,102 | +63% | 0 | 0 | — |
case-13 | pass→pass | 11,680 | 8,952 | -23% | 1 | 1 | 0% | 2,015 | 2,535 | +26% | 0 | 0 | — |
case-14 | pass→pass | 10,502 | 5,651 | -46% | 1 | 1 | 0% | 1,887 | 2,097 | +11% | 0 | 0 | — |
case-15 | pass→pass | 11,285 | 6,603 | -41% | 1 | 1 | 0% | 2,032 | 2,263 | +11% | 0 | 0 | — |
case-16 | pass→pass | 8,804 | 6,174 | -30% | 1 | 1 | 0% | 1,482 | 2,043 | +38% | 0 | 0 | — |
case-17 | fail→pass | 9,581 | 5,970 | -38% | 1 | 1 | 0% | 1,388 | 1,924 | +39% | 0 | 0 | — |
case-18 | pass→pass | 15,446 | 15,187 | -2% | 1 | 1 | 0% | 2,832 | 3,640 | +29% | 0 | 0 | — |
case-19 | pass→pass | 18,431 | 14,863 | -19% | 1 | 1 | 0% | 2,924 | 3,582 | +23% | 0 | 0 | — |
case-20 | pass→pass | 13,857 | 12,445 | -10% | 1 | 1 | 0% | 2,723 | 3,652 | +34% | 0 | 0 | — |
case-21 | pass→pass | 13,110 | 8,795 | -33% | 1 | 1 | 0% | 2,423 | 2,760 | +14% | 0 | 0 | — |
case-22 | pass→pass | 11,019 | 6,336 | -42% | 1 | 1 | 0% | 2,068 | 2,166 | +5% | 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.