Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

/learn - Hizli Ogrenim Kaydi — measured AI agent skill | DecimalAI

vibeeval/learn

NewSkillScoreMaintained

Hizlica kural/ogrenim kaydet. CLAUDE.md'ye ve memory'ye yazar. Kullanim: /learn <kural>

by vibeeval

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Knowledge Management pack

decimalai skills pull vibeeval-learn

Get this skill — curl · raw file

curl -fsSL https://app.decimal.ai/s/vibeeval-learn/SKILL.md

Raw file: /s/vibeeval-learn/SKILL.md ↗ · machine-readable summary at /s/vibeeval-learn.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

22 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 6d ago

—+35% tokens0% turns27% of whole cases pass with skill (6/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-09	✗→✓	▲ Improved	26%	0%
case-10	✗→✓	▲ Improved	39%	0%
case-02	✗→✓	▲ Improved	209%	0%
case-20	✓→✓	＝ Same ✓	25%	0%
case-21	✓→✓	＝ Same ✓	-2%	0%

Showing 5 of 22 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 8/1/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

vibeeval / vibecosystem

skills/learn/SKILL.md

View on GitHub ↗

License MITImported 8/1/2026Version v1

Skill contents

SKILL.md

/learn - Hizli Ogrenim Kaydi

Kullanici /learn <kural> yazdiginda su adimlari takip et:

Adim 1: Argumani Parse Et

Kullanicinin verdigi metni analiz et:

Ne ogrenilmeli?
Severity: CRITICAL / IMPORTANT / MINOR
Kategori: code / react / api / git / security / performance / testing / workflow

Adim 2: CLAUDE.md'ye Kaydet

Mevcut dizinde CLAUDE.md var mi kontrol et.

CLAUDE.md VARSA:

"LEARNED MISTAKES" bolumune ekle:

markdown
- [TARIH] HATA: <ogrenim> | COZUM: <ne yapilmali> | ONLEM: <kural>

"ERROR TRACKING" tablosuna ekle:

markdown
| TARIH | kategori | - | 1 | Learned | Yes |

CLAUDE.md YOKSA:

Kullaniciya sor: "Bu dizinde CLAUDE.md yok. Olusturayim mi? (template: ~/.claude/templates/CLAUDE-TEMPLATE.md)"

Adim 3: Memory'ye Kaydet

Genel bir ogrenim ise (sadece bu projeye ozel degilse), memory sistemine de kaydet:

bash
cd ~/.claude && PYTHONPATH=scripts python3 scripts/core/store_learning.py \
  --session-id "learn-command" \
  --content "<ogrenim>" \
  --context "<baglam>" \
  --tags "learn,<kategori>" \
  --confidence high

Adim 4: Onay Ver

OGRENILDI:
  Kural: <kural>
  Severity: CRITICAL/IMPORTANT/MINOR
  Kaydedildi: CLAUDE.md + memory
  Kategori: <kategori>

Ornekler

/learn API route'larda try-catch sart
-> CLAUDE.md'ye IMPORTANT/api olarak kaydeder

/learn .env dosyasini ASLA commit'leme
-> CLAUDE.md'ye CRITICAL/security olarak kaydeder

/learn React'ta useEffect cleanup unutma
-> CLAUDE.md'ye IMPORTANT/react olarak kaydeder

/learn Bu projede port 3737 kullaniliyor
-> CLAUDE.md'ye MINOR/workflow olarak kaydeder (memory'ye kaydetmez, proje-ozel)

Kurallar

Kisa ve net yaz - gereksiz uzatma
Her ogrenim actionable olmali (ne yapilmali acik)
Ornek kod ekle mumkunse
Tekrar kontrolu yap - ayni kural varsa "tekrar" sayisini artir
Tarih otomatik eklenir (bugunun tarihi)

Want to see if vibeeval/learn would activate on your phrasing? Try it before you install — no LLM call, no signup.

Other measured skills in the registry, with their headline benchmark lift.