pipeline-salvation

What problem does it solve?

pipeline-salvation helps you recover the working core of a messy, partially-understood, or mid-implementation CRM/ETL/data-pipeline without collapsing into a full rewrite or accidental behavior changes.

Core Features & Use Cases

• Confidence-first recovery methodology: define measurable success thresholds before touching code, and gate progress by evidence.
• Explicit execution boundaries: maintain clear stage contracts for both the Python layer and Workflowps/ETL engine, including done-sentinels and artifact expectations.
• Runnable spine + probe discipline: identify gaps precisely, wire new gates as warn-only initially, and promote to hard-block only after Level-4 staging evidence.
• Guardrail assessment & file bucketing: classify everything into KEEP / REWRITE-MINIMAL / DEFER / DROP so you only change what must be changed.

Quick Start

Use pipeline-salvation as the methodology reference before writing a plan for a messy pipeline, ensuring you set success thresholds, stage contracts, and probe/guardrail rules first.