Ready to supercharge your AI agents?

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

Get Started Free →

ETL Pipeline Builder Skill · a5c-ai | DecimalAI

a5c-ai/etl-pipeline-builder

SkillScore94Maintained

Build and manage ETL pipelines for data migration with transformation, CDC, and monitoring

Databy a5c-ai

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

Automatic

Data Engineer pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/a5c-ai-etl-pipeline-builder/SKILL.md

decimalai skills pull a5c-ai-etl-pipeline-builder --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/a5c-ai-etl-pipeline-builder/SKILL.md https://app.decimal.ai/s/a5c-ai-etl-pipeline-builder/SKILL.md

Raw file: /s/a5c-ai-etl-pipeline-builder/SKILL.md ↗ · machine-readable summary at /s/a5c-ai-etl-pipeline-builder.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

Runs

5,659

all versions · 30d

Eval pass

98%

global

Used when offered

100%

used / offered · all versions

By model· real traffic, every org · 30d — not the benchmark

Model	Eval pass	Runs
gemini-3.6-flashlowest	96%	50
gemini-3.1-pro-preview	100%	3

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 54d ago

Yes — +86% lift+10% tokens0% turns86% of whole cases pass with skill (19/22)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-01	✗→✓	▲ Improved	-14%	0%
case-02	✗→✓	▲ Improved	-32%	0%
case-03	✗→✓	▲ Improved	27%	0%
case-07	✗→✓	▲ Improved	-14%	0%
case-08	✗→✓	▲ Improved	-7%	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 — 7/31/2026

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

a5c-ai / babysitter

library/specializations/code-migration-modernization/skills/etl-pipeline-builder/SKILL.md

View on GitHub ↗

License MITImported 7/31/2026Version v1

Skill contents

SKILL.md

ETL Pipeline Builder Skill

Builds and manages ETL (Extract, Transform, Load) pipelines for data migration, supporting incremental loads, CDC, and comprehensive monitoring.

Purpose

Enable data pipeline creation for:

Source-to-target mapping
Transformation definition
Incremental load setup
CDC configuration
Pipeline monitoring

Capabilities

1. Source-to-Target Mapping

Define column mappings
Handle schema differences
Configure data type conversions
Manage derived columns

2. Transformation Definition

Data type transformations
Value mappings
Aggregations
Lookups and enrichments

3. Incremental Load Setup

Define watermarks
Configure incremental columns
Handle deletes
Manage merge logic

4. CDC Configuration

Log-based CDC
Trigger-based CDC
Timestamp-based CDC
Full load comparison

5. Error Handling

Define retry policies
Configure dead letter queues
Handle data quality issues
Implement alerting

6. Pipeline Monitoring

Track pipeline metrics
Monitor data volumes
Alert on failures
Generate SLA reports

Tool Integrations

| Tool | Type | Integration Method | |------|------|-------------------| | Apache Airflow | Orchestration | Python | | dbt | Transformation | CLI | | Airbyte | Data integration | API | | Fivetran | SaaS ETL | API | | AWS DMS | Cloud migration | CLI | | Debezium | CDC | Config |

Output Schema

json
{
  "pipelineId": "string",
  "timestamp": "ISO8601",
  "pipeline": {
    "name": "string",
    "source": {},
    "target": {},
    "mappings": [],
    "transformations": [],
    "schedule": "string"
  },
  "artifacts": {
    "dagFile": "string",
    "configFile": "string",
    "sqlFiles": []
  },
  "deployment": {
    "status": "string",
    "url": "string"
  }
}

Integration with Migration Processes

database-schema-migration: Data movement
cloud-migration: Cloud data pipelines
data-format-migration: Format transformation

Related Skills

data-migration-validator: Validation
schema-comparator: Schema mapping

Related Agents

database-migration-orchestrator: Pipeline orchestration
data-architect-agent: Pipeline design

Illustrative Want to see what a5c-ai/etl-pipeline-builder would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

8/1/2026, 01:52 UTC · 22 cases

Lift

+86%

86% vs 0%

Agent turns

1.0 vs 1.0

Tokens

+10%

3,232.82 vs 2,939.45

fail → pass · 19fail → fail · 3

Per-case results (22)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-01 I need to set up an automated data movement pipeline from our production PostgreSQL database to Snowflake using Airflow. We want to capture recent updates based on modified timestamps and transform user attributes into a unified reporting schema. Please provide the JSON output containing the pipeline specification, source-target mappings, generated Airflow DAG code, and deployment status details.	fail→pass	18,892	15,055	-20%	1	1	0%	5,157	4,421	-14%	0	0	—
case-02 We are migrating legacy MySQL order records into an AWS Redshift warehouse and need continuous log-based change capture configured. The process should sanitize missing fields and generate aggregate daily sales metrics. Generate a complete JSON response structured with the pipeline identifier, schema mapping rules, generated SQL transformation scripts, orchestration DAG files, and deployment links.	fail→pass	32,470	17,863	-45%	1	1	0%	6,205	4,197	-32%	0	0	—
case-03 Help me construct an ETL workflow to ingest operational metrics from a REST API target into BigQuery with retry policies and dead-letter routing for bad records. I need the response formatted as a structured JSON object containing pipeline metadata, transformation logic, configuration artifacts, and deployment status.	fail→pass	22,921	17,967	-22%	1	1	0%	4,112	5,204	+27%	0	0	—
case-04 Write PostgreSQL DDL statements to add a foreign key constraint named fk_orders_customers on orders(customer_id) referencing customers(id) and create a B-tree index named idx_orders_customer_id on orders(customer_id).	fail→fail	2,408	3,973	+65%	1	1	0%	504	1,210	+140%	0	0	—
case-05 Write an AWS IAM policy JSON document granting s3:GetObject access to the S3 bucket path arn:aws:s3:::corporate-data-lake/*.	fail→fail	2,876	3,869	+35%	1	1	0%	537	1,276	+138%	0	0	—
case-06 Provide the Python dictionary syntax for Airflow default_args setting retries to 3, retry_delay to 5 minutes, and execution_timeout to 1 hour.	fail→fail	2,832	2,398	-15%	1	1	0%	465	1,008	+117%	0	0	—
case-07 Design an incremental ETL pipeline from Oracle DB to AWS S3 using a high-watermark timestamp column updated_at. Return the response in standard JSON format containing pipeline metadata, mappings, transformations, schedule, dagFile, configFile, sqlFiles, deployment status and URL.	fail→pass	18,422	14,319	-22%	1	1	0%	4,144	3,568	-14%	0	0	—
case-08 Configure an ETL pipeline specification using dbt for SQL transformations on Snowflake data. Format the response as JSON containing pipelineId, timestamp, pipeline design, artifacts, and deployment status.	fail→pass	14,516	11,363	-22%	1	1	0%	2,942	2,737	-7%	0	0	—
case-09 Set up a log-based CDC pipeline pulling binary log changes from MySQL into Kafka. Output the structured JSON schema with pipelineId, timestamp, pipeline, artifacts, and deployment.	fail→pass	13,978	10,535	-25%	1	1	0%	2,713	2,736	+1%	0	0	—
case-10 Build a data ingestion pipeline that routes malformed CSV records from S3 into a dead letter queue while loading valid rows to Postgres. Output the pipeline definition JSON.	fail→pass	15,103	14,749	-2%	1	1	0%	3,227	3,711	+15%	0	0	—
case-11 Generate an ETL workflow definition for daily batch aggregation of analytics data in BigQuery using Airflow. Structure the answer using the pipeline JSON schema.	fail→pass	16,081	13,499	-16%	1	1	0%	3,555	3,248	-9%	0	0	—
case-12 Define a SaaS ETL synchronization pipeline for Salesforce leads into Redshift. Provide the output in the standard pipeline JSON structure.	fail→pass	12,036	10,938	-9%	1	1	0%	2,440	3,047	+25%	0	0	—
case-13 Configure an Airbyte API data integration pipeline moving DynamoDB data to Postgres. Format the response strictly using the output JSON schema.	fail→pass	10,865	14,532	+34%	1	1	0%	2,201	4,040	+84%	0	0	—
case-14 Specify a cloud database migration pipeline from on-premise SQL Server to AWS Aurora PostgreSQL using AWS DMS. Return the output as JSON.	fail→pass	16,515	11,211	-32%	1	1	0%	3,311	3,291	-1%	0	0	—
case-15 Design an ETL pipeline that calculates a derived full_name column from first_name and last_name during ingestion from MongoDB to Databricks. Express the result as JSON.	fail→pass	10,603	10,437	-2%	1	1	0%	2,308	2,919	+26%	0	0	—
case-16 Configure a trigger-based CDC pipeline capturing inserts and updates on a Postgres audit table into ClickHouse. Structure the response using the standard JSON specification.	fail→pass	12,151	15,721	+29%	1	1	0%	2,477	4,121	+66%	0	0	—
case-17 Create an ETL pipeline converting string dates to epoch timestamps and string amounts to numeric decimals during CSV to Parquet loading. Return the structured JSON output.	fail→pass	13,249	11,475	-13%	1	1	0%	2,573	3,216	+25%	0	0	—
case-18 Build an ETL pipeline specification that includes SLA monitoring metrics and volume tracking for daily clickstream data loading into Redshift. Format as JSON.	fail→pass	12,006	15,831	+32%	1	1	0%	2,463	3,947	+60%	0	0	—
case-19 Design an ETL pipeline that enriches incoming transaction events with customer geography lookups before loading into Snowflake. Output the pipeline JSON object.	fail→pass	15,566	9,721	-38%	1	1	0%	3,584	2,734	-24%	0	0	—
case-20 Configure an ETL pipeline that detects changes between legacy nightly exports and target data warehouse using full load comparison. Output the formatted JSON result.	fail→pass	15,789	9,076	-43%	1	1	0%	3,357	2,591	-23%	0	0	—
case-21 Construct an ETL pipeline mapping legacy XML product catalogs to a JSON relational model in BigQuery with schema mapping and type conversions. Format response as JSON.	fail→pass	16,324	14,442	-12%	1	1	0%	3,617	3,808	+5%	0	0	—
case-22 Define an ETL pipeline aggregating telemetry logs from multiple S3 buckets into weekly summary tables in Redshift. Produce the output using the required JSON schema.	fail→pass	13,356	14,675	+10%	1	1	0%	2,776	4,092	+47%	0	0	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

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 +86 percentage points is the difference between those two pass rates over the 22 comparable cases.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 0%
Pass rate with the skill: 86.4%
Cases attempted: 22
Cases counted toward lift: 22
Cases that regressed: 0
Skill version measured: v1
Run completed: 8/1/2026, 01:52 UTC

Measured example — same prompt, with vs. without the skill · from the benchmark run · case case-01

Prompt

I need to set up an automated data movement pipeline from our production PostgreSQL database to Snowflake using Airflow. We want to capture recent updates based on modified timestamps and transform user attributes into a unified reporting schema. Please provide the JSON output containing the pipeline specification, source-target mappings, generated Airflow DAG code, and deployment status details.

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.

Show both answers

SkillScore blends 4 signals3 of 4 complete

BenchmarkPass rate WITH skill on a verified run86

Live evalPass rate on real evaluated traces98%

AI ratingNeeds ≥10 rated traces

AdoptionBreadth of real use across orgs100

Backed by 3 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 54 days ago

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