Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Agent ve skill dosyalarinin yapisal dogrulamasi. Frontmatter kontrol, naming convention, zorunlu bolum kontrolu, tutarlilik denetimi. Yeni agent/skill eklendiginde veya mevcut dosyalar duzenlediginde otomatik calistirilir.
.claude/skills/agent-linter/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-19 | ✗→✓ | ▲ Improved | — | — |
| case-21 | ✗→✓ | ▲ Improved | — | — |
| case-13 | ✗→✓ | ▲ Improved | — | — |
| case-11 | ✗→✓ | ▲ Improved | — | — |
| case-08 | ✗→✓ | ▲ Improved | — | — |
Agent ve skill dosyalarinin kalite kontrolu. Yanlis yapilandirilmis dosyalar runtime'da sessizce basarisiz olur -- linter bunu onler.
yaml--- name: agent-adi # kebab-case, dosya adiyla AYNI olmali description: Tek satir # 20-200 karakter arasi tools: ["Read", "Bash"] # Gecerli tool listesi model: opus # Opsiyonel: opus | sonnet ---
| Kural | Kontrol | Hata Seviyesi | |-------|---------|---------------| | name mevcut | Frontmatter'da name var mi? | ERROR | | name = dosya adi | name: sleuth == sleuth.md | ERROR | | name kebab-case | code-reviewer OK, codeReviewer FAIL | ERROR | | description mevcut | Bos olmamali | ERROR | | description uzunlugu | 20-200 karakter | WARN | | tools array | String array olmali | ERROR | | tools gecerli | Sadece bilinen tool isimleri | WARN | | model gecerli | opus, sonnet veya bos | ERROR |
| Kural | Kontrol | Hata Seviyesi | |-------|---------|---------------| | System prompt var | Frontmatter'dan sonra icerik var | ERROR | | 50+ satir | Cok kisa agent tanimlari yetersiz | WARN | | 2000 satir max | Cok uzun dosyalar token israf eder | WARN | | Markdown basliklar | En az 1 ## basligi var | WARN |
yaml--- name: skill-adi # kebab-case description: Aciklama # 20-300 karakter ---
| Kural | Kontrol | Hata Seviyesi | |-------|---------|---------------| | Dosya adi | SKILL.md veya skill.md | ERROR | | name mevcut | Frontmatter'da | ERROR | | name = klasor adi | name: tdd-workflow == skills/tdd-workflow/ | WARN | | description mevcut | Bos olmamali | ERROR | | Icerik var | Frontmatter'dan sonra en az 20 satir | WARN | | Kod ornegi var | En az 1 code block | INFO |
bash#!/bin/bash # scripts/lint-agents.sh ERRORS=0 WARNS=0 lint_agent() { local file="$1" local name=$(basename "$file" .md) # Frontmatter kontrolu if ! head -1 "$file" | grep -q "^---"; then echo "ERROR: $file - frontmatter yok" ERRORS=$((ERRORS + 1)) return fi # name field kontrolu local fm_name=$(sed -n '/^---$/,/^---$/p' "$file" | grep "^name:" | sed 's/name: *//') if [ -z "$fm_name" ]; then echo "ERROR: $file - name field yok" ERRORS=$((ERRORS + 1)) elif [ "$fm_name" != "$name" ]; then echo "ERROR: $file - name '$fm_name' != dosya adi '$name'" ERRORS=$((ERRORS + 1)) fi # description kontrolu local desc=$(sed -n '/^---$/,/^---$/p' "$file" | grep "^description:") if [ -z "$desc" ]; then echo "ERROR: $file - description yok" ERRORS=$((ERRORS + 1)) fi # Uzunluk kontrolu local lines=$(wc -l < "$file") if [ "$lines" -lt 10 ]; then echo "WARN: $file - cok kisa ($lines satir)" WARNS=$((WARNS + 1)) elif [ "$lines" -gt 2000 ]; then echo "WARN: $file - cok uzun ($lines satir)" WARNS=$((WARNS + 1)) fi } echo "=== Agent Linter ===" for f in agents/*.md; do lint_agent "$f" done echo "" echo "=== Skill Linter ===" for f in skills/*/SKILL.md skills/*/skill.md; do [ -f "$f" ] && lint_agent "$f" done echo "" echo "Sonuc: $ERRORS error, $WARNS warning" [ "$ERRORS" -gt 0 ] && exit 1 || exit 0
yaml# .github/workflows/lint-agents.yml name: Agent & Skill Lint on: pull_request: paths: ['agents/**', 'skills/**'] jobs: lint: runs-on: ubuntu-latest steps: - uses: actions/checkout@v4 - run: bash scripts/lint-agents.sh
| Hata | Ornek | Cozum | |------|-------|-------| | name/dosya uyumsuzlugu | name: codeReviewer dosya: code-reviewer.md | name'i kebab-case yap | | Description cok kisa | description: Reviews code | Ne yaptigini detayla | | Tools yanlis format | tools: Read, Bash | tools: ["Read", "Bash"] | | Bos agent dosyasi | Sadece frontmatter, body yok | System prompt ekle | | Duplicate name | 2 agent ayni name'e sahip | Benzersiz isim ver | | Skill klasor/name uyumsuz | name: tdd klasor: tdd-workflow/ | Eslestir |
bash# Tum agent'lari tara, rapor uret cd ~/vibecosystem bash scripts/lint-agents.sh 2>&1 | tee lint-report.txt # Sadece ERROR'lari goster grep "^ERROR" lint-report.txt # Istatistik echo "Agents: $(ls agents/*.md | wc -l)" echo "Skills: $(ls skills/*/SKILL.md skills/*/skill.md 2>/dev/null | wc -l)" echo "Errors: $(grep -c ERROR lint-report.txt)" echo "Warnings: $(grep -c WARN lint-report.txt)"
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-16 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-15 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | pass→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
case-23 | fail→pass | — | — | — | — | — | — | — | — | — | — | — | — |
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. 23 cases were attempted, and 22 counted toward the lift figure. The other 1 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 +43 percentage points is the difference between those two pass rates over the 22 comparable cases.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
Other measured skills in the registry, with their headline benchmark lift.