Die Oberfläche ist lokalisiert; Skill-Beschreibungen und -Dokumente werden per KI automatisch in Ihre Sprache übersetzt. Code, Befehle, Pfade und Fachbegriffe bleiben im Original, maßgeblich ist das Dokument des Autors.
Skills entdecken
Daten und Forschung · SKILL.md

sql-connectors

sql-connectors

Guides awswrangler Redshift, Data API, and direct MySQL, PostgreSQL, SQL Server, and Oracle connector workflows.

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill-Dokumentation

GitHub

name: sql-connectors description: "Guides awswrangler Redshift, Data API, and direct MySQL, PostgreSQL, SQL Server, and Oracle connector workflows." disable-model-invocation: true metadata: disco-role: operating license: Apache 2.0

SQL Connectors

Use this sub-skill when the user is working with relational database connections or SQL copy/unload flows.

Typical triggers:

  • "connect to Redshift"
  • "read SQL from MySQL or PostgreSQL"
  • "write a pandas DataFrame to SQL Server"
  • "use the RDS Data API"
  • "use Redshift COPY or UNLOAD"
  • "open an Oracle connection"

Read references/api-reference.md for the connector families and their key methods. Read references/workflows.md for choosing the right connection path and moving data in or out. Read references/troubleshooting.md for missing driver, secret, timeout, SSL, and permission failures. Use ../../scripts/check_runtime.py from the root skill to confirm the required connector extras before attempting a live database workflow.

Include here

  • Redshift direct connector workflows, including connect_temp, copy, and unload.
  • Redshift and RDS Data API workflows.
  • MySQL, PostgreSQL, SQL Server, and Oracle direct connectors.
  • read_sql_query, read_sql_table, and to_sql for the supported database families.

Exclude or route elsewhere

  • S3 file staging, dataset reads and writes, and Delta Lake go to s3-lakehouse.
  • Glue Catalog, Athena, Data Quality, and Clean Rooms go to catalog-and-query.
  • Runtime, extras, and Ray/Modin mode questions go to runtime-and-config.
  • DynamoDB, Timestream, OpenSearch, Neptune, QuickSight, CloudWatch, EMR, STS, Secrets Manager, and Chime go to service-integrations.

Common user outcomes

A future agent should be able to:

  • choose the correct connector family,
  • open a connection with the right secret, catalog, or database arguments,
  • run parameterized SQL,
  • write pandas data back to a table,
  • and decide when Redshift COPY/UNLOAD or Data API is the better fit.

Shared references

  • references/api-reference.md
  • references/workflows.md
  • references/troubleshooting.md
Gespeicherte Skills