사용 방법
- 설치 명령을 복사하고 자신의 에이전트 환경에 설치하기 전에 소스 패키지를 확인하세요.
- 필요한 도구, 인증 정보, 지원 환경은 원본 문서에서 확인하세요.
- 입력과 원하는 산출물을 설명하고 출처와 결과를 대조하세요.
npx skills add https://github.com/majesticlabs-dev/majestic-marketplace --skill etl-patterns시작 전 확인
- 설치에는 모델 구독, API 한도, 파일 접근 권한이 포함되지 않습니다.
- 별점 수와 설치 수는 활동 지표이며 독립적인 품질·보안 검증이 아닙니다.
원본 Skill 문서
원본 저장소의 제목, 예시, 코드, 표, 링크, 이미지를 유지해 표시합니다.
ETL Patterns
Orchestrator for production-grade Extract-Transform-Load patterns.
Skill Routing
| Need | Skill | Content |
|---|---|---|
| Reliability patterns | etl-core-patterns | Idempotency, checkpointing, error handling, chunking, retry, logging |
| Load strategies | etl-incremental-patterns | Backfill, timestamp-based, CDC, pipeline orchestration |
Pattern Selection Guide
By Reliability Need
| Need | Pattern | Skill |
|---|---|---|
| Repeatable runs | Idempotency | etl-core-patterns |
| Resume after failure | Checkpointing | etl-core-patterns |
| Handle bad records | Error handling + DLQ | etl-core-patterns |
| Memory management | Chunked processing | etl-core-patterns |
| Network resilience | Retry with backoff | etl-core-patterns |
| Observability | Structured logging | etl-core-patterns |
By Load Strategy
| Scenario | Pattern | Skill |
|---|---|---|
| Small tables (<100K) | Full refresh | etl-incremental-patterns |
| Large tables | Timestamp incremental | etl-incremental-patterns |
| Real-time sync | CDC events | etl-incremental-patterns |
| Historical migration | Parallel backfill | etl-incremental-patterns |
| Zero-downtime refresh | Swap pattern | etl-incremental-patterns |
| Multi-step pipelines | Pipeline orchestration | etl-incremental-patterns |
Quick Reference
Idempotency Options
python
# Small datasets: Delete-then-insert
# Large datasets: UPSERT on conflict
# Change detection: Row hash comparisonLoad Strategy Decision
Is table < 100K rows?
→ Full refresh
Has reliable timestamp column?
→ Timestamp incremental
Source supports CDC?
→ CDC event processing
Need zero downtime?
→ Swap pattern (temp table → rename)
One-time historical load?
→ Parallel backfill with date rangesCommon Pipeline Structure
python
# 1. Setup
checkpoint = Checkpoint('.etl_checkpoint.json')
processor = ETLProcessor()
# 2. Extract (with incremental)
df = incremental_by_timestamp(source_table, 'updated_at')
# 3. Transform (with error handling)
transformed = processor.process_batch(df.to_dict('records'))
# 4. Load (with idempotency)
upsert_records(pd.DataFrame(transformed))
# 5. Checkpoint
checkpoint.set_last_processed('sync', df['updated_at'].max())
# 6. Handle failures
processor.save_failures('failures/')Related Skills
data-validation- Validate data quality during ETLdata-quality- Monitor data quality metricspandas-coder- DataFrame transformations
