Repeatability
High
The task runs weekly with the same three sources and the same output schema every time. Column structures may vary slightly across platforms but are stable within each platform, making this highly automatable once the mapping is defined.
Ambiguity Tolerance
High
Success criteria are explicit: unique attendees by ticket_id and email, standardized dates, platform attribution, and a clean CSV output. There is little room for interpretation about what 'done' looks like.
Data & Tool Availability
High
The agent only needs access to the three CSV exports, which the user already produces manually. No live API access or authentication is required — just file input and a script or data agent to process them.
Error Cost
Low
The output is an internal reporting artifact, not a customer-facing or financial action. Errors are easily caught by spot-checking row counts and are fully reversible by re-running the merge with corrected logic.
Human Judgment Required
Low
Deduplication rules, date normalization, and column mapping are all rule-based. The only edge case requiring judgment is conflicting records for the same attendee across platforms, which can be handled with a documented tie-breaking rule.