data-pipeline-reliability

Installation
SKILL.md

Data Pipeline Reliability

Use this skill when the main question is whether a pipeline can be rerun, retried, backfilled, or recovered without corrupting data.

Boundaries

  • Use data-engineering when the main decision is platform shape, marts, semantic metrics, or serving layers.
  • Use data-architecture-strategy when the main decision is which architecture family to adopt.
  • Use data-quality when the main concern is whether outputs are trustworthy once they are produced.

Default procedure

  1. Identify the pipeline mode:
    • full refresh
    • incremental batch
    • CDC or streaming
    • backfill or replay
  2. Define the failure mode to guard against:
    • duplicate writes
Related skills

More from jimnguyendev/jimmy-skills

Installs
4
GitHub Stars
4
First Seen
Apr 23, 2026