Audit Claude Code agents, skills, and commands for quality and production readiness. Use when evaluating skill quality, checking production readiness scores, or comparing agents against best-practice templates.
日本語の概要は準備中です。原文の説明を表示しています。
Push current branch and return the pipeline tracking URL (GitLab or GitHub Actions)
インストールする前に、エージェントに与えられる指示の中身を確認できます。
Pushes the current branch and returns the pipeline tracking link.
BRANCH=$(git branch --show-current)
# 1. Safety checks
if echo "$BRANCH" | grep -qE "^(main|master|production)$"; then
echo "❌ Direct push to $BRANCH is not allowed. Create a feature/fix branch."
exit 1
fi
# 2. Check for uncommitted changes
UNCOMMITTED=$(git status --porcelain | wc -l | tr -d ' ')
if [ "$UNCOMMITTED" -gt 0 ]; then
echo "⚠️ $UNCOMMITTED uncommitted file(s):"
git status --short
echo ""
echo "Commit first with /commit or git add + git commit"
exit 0
fi
# 3. Push
echo "Pushing → origin/$BRANCH"
git push origin "$BRANCH" 2>&1
REMOTE=$(git remote get-url origin 2>/dev/null || echo "")
WEB_URL=$(echo "$REMOTE" | sed 's/git@gitlab\.com:/https:\/\/gitlab.com\//' | sed 's/\.git$//')
echo ""
echo "✅ Push OK"
echo " Pipeline: $WEB_URL/-/pipelines?ref=$BRANCH"
# Live status via glab (if installed)
if command -v glab &>/dev/null; then
sleep 3
glab ci status --branch "$BRANCH" 2>/dev/null || true
fi
REMOTE=$(git remote get-url origin 2>/dev/null || echo "")
WEB_URL=$(echo "$REMOTE" | sed 's/git@github\.com:/https:\/\/github.com\//' | sed 's/\.git$//')
echo ""
echo "✅ Push OK"
echo " Actions: $WEB_URL/actions?query=branch%3A$BRANCH"
# Live status via gh (if installed)
if command -v gh &>/dev/null; then
sleep 5
gh run list --branch "$BRANCH" --limit 3 2>/dev/null || true
fi
Pushing → origin/feat/add-payment-retry
Enumerating objects: 12, done.
...
✅ Push OK
Pipeline: https://gitlab.com/org/my-app/-/pipelines?ref=feat/add-payment-retry
Pipeline status (after 3s):
⏳ lint running
⏸️ test waiting
⏸️ deploy waiting
/ci:pipeline
Target: $ARGUMENTS
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Audit Claude Code agents, skills, and commands for quality and production readiness. Use when evaluating skill quality, checking production readiness scores, or comparing agents against best-practice templates.
日本語の概要は準備中です。原文の説明を表示しています。
Codebase health audit scoring 7 categories with progression plan
日本語の概要は準備中です。原文の説明を表示しています。
Autonomous improvement loop: scan codebase metrics, scaffold experiment files, run agent-driven iterations until metric improves
日本語の概要は準備中です。原文の説明を表示しています。
Generate bounded independent candidates, score them against a frozen rubric, and verify the selected result with a proof log.
日本語の概要は準備中です。原文の説明を表示しています。
Post-deploy monitoring: watch production after a deploy and alert on regressions
日本語の概要は準備中です。原文の説明を表示しています。
Restore context after /clear by summarizing recent work and project state
日本語の概要は準備中です。原文の説明を表示しています。