Third-party data importers
Key every upsert on the provider's external id and advance a stored cursor — then re-running a sync is safe, not a duplicate machine.
What this skill helps you build
Sync data from an external API into your own schema — paginated pulls, idempotent upserts by external id, and cursor-based incremental re-sync.
The production takeaway
Key every upsert on the provider's external id and advance a stored cursor — then re-running a sync is safe, not a duplicate machine.
Inside this skill
The full skill expands these implementation areas with decisions, edge cases, prompts, tests, and framework-specific code.
Unlock the full implementation
Paid access includes the complete skill body, implementation prompt, common mistakes, production checklist, and code examples where this skill includes them.