Repeatability
High
The transformation rules are fixed and structural: date normalization, location code mapping, cost standardization, and flag logic are all deterministic. This task is highly repeatable and could be scripted once and rerun on future exports.
Ambiguity Tolerance
High
Success criteria are explicit and enumerable: ISO dates, per-unit costs, normalized location codes, blank-cost rows removed, zero-quantity stale SKUs flagged. An agent can verify each rule independently with no subjective judgment required.
Data & Tool Availability
Medium
The main CSV is available, but the footnote sheet containing case-vs-unit mappings is a separate file that must also be provided and must be machine-readable. If the footnotes are unstructured prose or inconsistently formatted, the agent will need human help to parse the conversion logic.
Error Cost
Medium
Incorrect cost normalization could propagate bad unit costs into purchasing or pricing decisions, which has real financial consequences. However, the output is a CSV that a human can audit before ingestion, and the source data is preserved, making errors detectable and reversible.
Human Judgment Required
Low
No taste, ethics, or relationship context is needed. The only judgment call is interpreting ambiguous footnotes, which is a bounded parsing problem rather than a subjective one — and can be resolved by providing a clean mapping table upfront.